Why Warren Buffett Sells Bank Stocks (And What He's Buying Instead)

Short answer: Warren Buffett sells bank stocks because the math turned against them, not because banks are about to collapse. He sold roughly half of his Bank of America stake — about 515.6 million shares — between July 2024 and December 2025¹. He exited Citigroup completely in the first quarter of 2025². Wells Fargo, JPMorgan, Goldman Sachs, and a handful of other banks were already gone years earlier³. The reason is simple: he bought banks cheap, the prices ran up, and the trade stopped working. He has not warned of a banking crisis. He is just doing what he has always done.

If you own a bank stock and you read the news that Warren Buffett sells bank stocks again, you panic. That panic is the wrong response. Buffett is the most-watched investor in the world. His moves get described as oracles. But the reasons behind those moves are usually mundane — a price he paid years ago no longer makes sense at today's quote. Once you see the math, the headlines lose their drama.

This guide is the honest answer in plain English. We will lay out the exact tally of every bank Buffett sold and roughly when. We will walk through the three boring reasons behind the selling. Then we will tell you what it actually means for a retail investor holding a bank stock today. No finance degree needed.

The exact tally of every bank Warren Buffett sold

People talk about Warren Buffett sells bank stocks like it is one decision. It is not. It is fifteen years of slow, deliberate exits. Here is the actual list, in roughly the order it happened.

• Goldman Sachs (GS) — bought during the 2008 crisis on preferred-stock terms, exited completely by 2020³.
• JPMorgan Chase (JPM) — bought in 2018, fully exited by the end of 2020³. He held it for two years.
• M&T Bank (MTB) — exited 2020³.
• PNC Financial (PNC) — exited 2021³.
• U.S. Bancorp (USB) — long-time holding, exited across 2021 and 2022³.
• Wells Fargo (WFC) — Berkshire's largest bank position for decades. Fully exited by 2022³.
• Bank of New York Mellon (BK) — exited 2022³.
• Citigroup (C) — opened in 2022, fully exited in the first quarter of 2025 (the final 14.6 million shares, about $1 billion)².
• Capital One (COF) — held, but trimmed by 300,000 shares in Q1 2025, leaving 7.15 million².
• Bank of America (BAC) — about 515.6 million shares sold between July 2024 and December 2025, roughly half of his peak stake¹. Still a top-5 Berkshire holding⁴.

By the time Buffett handed the CEO chair to Greg Abel on January 1, 2026⁴, one major US bank position remained — Bank of America, at roughly half its peak size. That is a real reduction. It is also, when you look at the numbers, less dramatic than the headlines suggest. Berkshire still holds a stake worth tens of billions.

Why Warren Buffett sells bank stocks: three boring reasons

The financial press loves to assign drama to every Buffett trade. The truth is much quieter. There are three reasons he has been selling, and none of them are "the banks are about to fail".

Reason 1: He bought banks cheap, and they got expensive

This is the most important reason and the one you will rarely see in the headlines. Buffett bought his Bank of America preferred stake in August 2011, in the middle of a euro-zone scare. The common stock was trading at a 62 percent discount to its own book value at the time⁵. He was buying $1 of bank for less than 40 cents.

By early 2026, Bank of America's common shares traded at a 43 percent premium to book value⁵. The same business that cost 38 cents on the dollar in 2011 now cost $1.43 on the dollar. The trade Buffett opened was a cheap-bank trade. Once the cheapness was gone, the trade was gone with it.

This is the Buffett playbook. Buy when the math is obviously favorable. Sell when the math no longer is. It is not personal to the bank. It is the math.

Reason 2: Falling interest rates squeeze bank earnings

Big US banks make money in two main ways. They charge borrowers more than they pay savers (the net interest margin, or the spread). And they collect fees on cards, trading, and wealth management. The first part — the spread — depends heavily on where short-term and long-term interest rates sit relative to each other.

When the Federal Reserve cuts rates aggressively, the gap between what banks earn on loans and what they pay on deposits compresses. Net interest income falls. For a bank like Bank of America, whose balance sheet is unusually exposed to falling short rates, the hit is real⁶. Buffett does not need to predict rates to act on this. He just needs to notice that the easy upside from the post-2022 rate-hike cycle is over.

Reason 3: Taxes are unusually low — so locking in gains makes sense

At Berkshire's May 2024 shareholder meeting, Buffett openly said he expected corporate tax rates to rise from current levels in the years ahead⁷. Federal corporate income tax in 2026 sits at 21 percent, one of the lowest rates in modern US history. Buffett's gains in Bank of America had grown into one of the largest unrealised positions on the books — billions of dollars in built-up capital gains.

If you believe taxes are going up, you sell now and pay 21 percent. If you wait and the rate climbs to, say, 28 percent, every dollar of gain you take later is worth 7 cents less. On a stake the size of Berkshire's Bank of America position, that difference adds up to billions. The selling pattern — heavy, steady, multi-quarter — is exactly what a tax-aware seller does. He is not panicking. He is realising gains while the rate is friendly.

What Buffett is buying instead of bank stocks

A Buffett sale is only half the story. The other half is where the money goes. The pattern over the last two years is consistent and revealing.

Chevron (CVX). Berkshire bought roughly 8.09 million shares of Chevron for about $1.2 billion in the fourth quarter of 2025⁸. Chevron has joined Bank of America in Berkshire's top-five holdings⁴. Big oil pays out real cash today. It does not depend on the Fed.

Domino's Pizza (DPZ). Four consecutive quarters of buying through 2025¹. Domino's is a high-return-on-capital franchise model. The economics are different from a bank in every way — no leverage, no interest-rate exposure, very simple unit math.

Constellation Brands (STZ). Berkshire doubled its stake in Q1 2025². Branded consumer staples. Long-life cash flows. Not a bank.

Alphabet (GOOGL). Under Abel's first full quarter, Berkshire tripled its existing stake in Q1 2026⁴. A different kind of business altogether — software margins, regulatory tailwinds and headwinds both.

Cash and short-term Treasuries. This is the biggest "buy" of all. By late 2024, Berkshire's cash and Treasury-bill pile crossed $300 billion⁹. Treasuries at 4–5 percent yield, with zero credit risk, are a serious competitor to any equity right now — including a bank stock paying a 2 percent dividend that depends on the rate cycle staying favorable. Buffett has not stopped investing. He has just decided cash is currently the better trade.

Notice what is not on the list. He has not bought a different bank. He has not rotated from Bank of America into a smaller, cheaper regional bank. The sale was not a bank-to-bank reshuffle. It was banks-to-elsewhere. That is the more meaningful signal than any single trade.

What Warren Buffett selling bank stocks means for you

You came here for the bottom line. Here it is.

If you own a US bank stock today and you read the news that Warren Buffett sells bank stocks, do not sell on the headline alone. Buffett's reasons are mostly about his specific tax position, his cost basis, and his alternative options (a $300 billion cash pile that earns a real yield). None of those three things apply to a retail investor in the same way. You probably bought your bank stock at a different price, in a different tax bracket, with a different alternative on the cash side.

The right question is not "is Buffett selling?". The right question is "does the math still work for me at today's price?". To answer that, you need three numbers:

1. The bank's price-to-book ratio today versus the average it has traded at over the last 10–15 years. If it is sitting above the high end of that range, you are paying a premium and the Buffett-style margin of safety is gone. We walk through what counts as a fair P/E and price-to-book in our what is a good P/E ratio guide.
2. The bank's return on tangible common equity (ROTCE) in the last reported quarter. Big banks should be doing high-teens or better in a normal environment. Below 10 percent suggests the business is being held back by either rates, credit losses, or both. The number sits in every quarterly 10-Q filed with the SEC¹⁰.
3. Your alternative. A 1-year Treasury bill is paying near 4 percent in mid-2026, federally tax-advantaged in many states, with zero credit risk. If your bank stock is paying a 2 percent dividend and the share price is flat-to-down, you are paying the bank stock to underperform a T-bill. Buffett spotted that asymmetry months before most retail holders did.

If the math still works at today's price for your situation, hold. If it does not, sell — but do it because of your own math, not because of his. The single most expensive mistake retail investors make is taking a famous investor's exit as a signal to copy without doing the work.

The Buffett lesson hidden in the bank sales

Underneath the bank trades, there is a Buffett lesson worth paying attention to. He has been doing this in public for over sixty years and the underlying behaviour barely changes.

He bought banks when banks were cheap. He held them while they were still cheap. He started trimming when they got fairly valued. He sold heavily when they became expensive and when his tax timing favoured the sale and when his alternative (Treasuries at 4–5 percent) became attractive on a risk-adjusted basis. Each individual reason on its own would not have been enough. All three together, on a position that had grown to be one of the largest in his portfolio, made the sale the obvious move.

This is the boring part of value investing that does not make for good television. The actual decision rule is price relative to value, given my alternatives, given my tax situation. It is arithmetic, applied patiently. There is no clever model, no insider tip, no proprietary indicator. Buffett's edge has always been that he actually does the work and waits.

If you are a retail investor and you want to do the same on your own portfolio, the recipe is identical. Ignore the headline. Look at the price-to-book and the return on equity. Compare it to a real alternative. Decide based on the math, not on what someone else did with a different cost basis and a different cash pile. That is the whole game.

What would change Buffett's mind on banks

For the curious — and for anyone watching the next Berkshire 13F filing — here is roughly what would have to happen for Buffett (or, now, Greg Abel) to start buying banks again.

• A US recession or credit shock that pushed major bank stocks back below book value. That is the cheap-bank trade resetting. Buffett re-opened Goldman, Bank of America, and others during exactly that kind of stress in 2008–2011. A repeat would likely look the same.
• A clear US interest-rate cycle reversal toward a higher long-term level. Banks make more on a steeper yield curve. If the Fed signalled a structurally higher long-end, the earnings power picture changes.
• A tax regime change that removed the incentive to lock in gains. If corporate tax rates drop materially below 21 percent, the urgency to sell unrealised gains evaporates.

None of those three things are visible in current data. So absent one of them showing up, expect the bank-selling pattern to continue under Greg Abel — though probably at a slower pace, since the easy gains have already been taken.

Related reading

For the broader question of how to know whether any stock is a buy, see our how to know if a stock is a good buy guide. For Buffett's own definition of what a fair price looks like in numbers, see what is a good P/E ratio. For the discounted-cash-flow approach that sits underneath every Buffett valuation, see our walkthrough on DCF valuation. For a closely-watched recent Buffett-style sale, see why Tesla stock is going down. For the original Buffett playbook in three pages, the margin of safety formula is the starting point.

For more plain-English essays on the value-investing way to look at a portfolio, see the rest of the Hub.