The world's biggest banks lent more to fossil fuels last year, not less

Climate & Energy 4 min 69 sources

65 banks pledged $906bn to coal, oil and gas in 2025 — an 8% rise — even as a US heat study warns of doubling hospital visits. The money and the damage are pulling in opposite directions.

Key takeaways

The world's 65 biggest banks lent $906bn to fossil fuels in 2025, up nearly 8% — the first year trackers expected a fall, and they got a rise instead.
The average rose, but 26 of the 65 banks actually cut their lending; the borrowers got funded anyway, because a loan one bank refuses, another writes.
A US study warns heat hospitalisations could double by 2040 — the money and the damage are pulling in opposite directions.

The banks went the wrong way

For the first time, the people tracking bank lending to fossil fuels expected the number to start falling. It rose instead.

The world’s 65 largest banks pledged $906bn to coal, oil and gas companies in 2025 — up $64bn, nearly 8% on the year before [9]. The figure comes from the annual Banking on Climate Chaos report, compiled by a coalition of environmental groups [9]. JPMorgan Chase led the list again, pushing $58bn to the sector, up 13% [9]. Bank of America came second, followed by the Japanese banks MUFG and Mizuho, then Citigroup. Barclays, at number eight, was the highest-placed British bank [9].

“Last year was the first year where we were hoping to see a continuous decrease,” said Caleb Schwartz, a policy analyst at one of the groups behind the report. “We actually saw that increase” [9]. Since the 2015 Paris agreement — where countries agreed to try to hold global heating to 1.5C above pre-industrial levels — the same banks have funnelled $8.7tn into digging and drilling [9]. Scientists now expect that 1.5C limit to be breached within years [9].

JPMorgan, asked about the lending, said it supports “the full range of energy solutions” and that its own data reflects its activity “more comprehensively and accurately than estimates by third parties” [9].

Why now, when the pressure should be the other way

Two things turned at once. First, oil and gas got more profitable. After the US and Israel’s strikes on Iran pushed up the global cost of oil and gas, several of the largest fossil fuel companies reported spiralling profits this year [9]. Lending follows return — money flows to where it pays, and right now fossil fuels pay.

Second, the political wind shifted. Several banks had announced targets to cut emissions and limit lending to the dirtiest fuels, like coal. Many have since walked those targets back amid political pressure in the US against climate commitments [9]. A pledge is cheap to drop when no one is enforcing it.

The lending is also getting more concentrated. A “dirty dozen” of banks now account for 40% of all fossil fuel funding, and almost all of it comes from just six places: the US, Canada, Japan, China, the UK and the EU [9]. The money for expanding existing oil and gas sites — not just maintaining them — jumped 27%, to $508bn [9].

Not everyone moved the same way

This is the part that matters. The average went up, but it hides a split. Twenty-six of the 65 banks actually reduced their fossil fuel lending last year [9]. The European banks BNP Paribas, UBS and La Caixa led the cuts [9].

So the headline isn’t “every bank doubled down.” It’s that the ones who pulled back were more than offset by the ones who leaned in. A loan one bank refuses, another writes. The borrower — Venture Global, Enbridge, Energy Transfer, the three biggest recipients in 2025 — gets the money either way [9].

The other column of the ledger

While the lending rose, a US study landed a number on what the burning produces. Heat-related emergency visits and hospitalisations in the US could roughly double by 2040 — from about 109,000 a year now to as many as 237,000 — pushing annual healthcare costs for heat illness above $1bn [5]. The study, in the journal GeoHealth, modelled 53 large metro areas [5]. Severe heat already kills more Americans each year than all other extreme weather combined, with deaths up more than 50% over two decades [5].

The damage won’t land evenly. Places not used to extreme heat — the north-east, the Ohio valley — are expected to suffer the worst from each major heat event, because they’re built and prepared for milder summers [5]. The people most at risk are the elderly, the sick, and those who can’t afford to run air conditioning [5]. The average US household is expected to spend nearly $800 on electricity this summer, up more than 10% on last year [5].

What’s quietly building anyway

None of this means the clean side has stalled. In Georgia, Qcells — the US arm of Korea’s Hanwha — started making solar cells this week at its Cartersville factory, a step toward building a full solar panel on US soil instead of importing the parts from Asia [3]. A federal judge restored a tax-credit route — the “5% safe harbor” — that wind and solar projects use to prove they qualify for clean-energy credits, after the administration tried to remove it [8]. And the Department of Energy reinstated a $57m grant to American Battery Technology, which will keep building a $115m lithium refinery on one of the largest known lithium deposits in the US [13].

These are small next to $906bn. But they’re the real economy moving — factories, refineries, rules — against the flow of the financing.

02 · Lesson · why it matters

When everyone wants the same thing and no one can be the one to do it

A loan one bank refuses, another writes — which is why a group can keep heading somewhere not one of them would choose.

The number that wasn’t supposed to go up

The people who count bank lending to fossil fuels expected last year to be the year it finally turned down. It went up instead — $906bn, nearly 8% more than the year before.

That’s strange if you think of the 65 banks as one decision-maker. No serious banker thinks the world is better off overheating. Many of these same banks made public promises to cut their fossil fuel lending. So how does a group of institutions, none of which wants a wrecked climate, keep funding the thing that wrecks it — and fund more of it?

The answer isn’t that they’re villains. It’s a trap built into the shape of the situation. And once you see the shape, you find it everywhere.

The trap has a name

Picture two banks. Both would prefer a stable climate. Both know the other one knows it. Now a profitable oil-expansion loan comes across the desk.

If JPMorgan turns it down on principle, the loan doesn’t vanish. Bank of America writes it. The oil gets drilled either way. JPMorgan loses the fee and the client — and the climate is no better off, because the project still happened. So the rational move for any single bank is: take the loan. The cost of restraint is yours alone; the benefit of restraint is shared with everyone, including the banks that didn’t restrain.

Run that logic across 65 banks and you get exactly what the report found. Twenty-six banks actually cut their lending last year. It didn’t matter. The ones who pulled back were more than covered by the ones who leaned in, and the borrowers — Venture Global, Enbridge, Energy Transfer — got funded anyway. The group walked toward an outcome not one member would have chosen for the group.

Why the average lies

There’s a quieter lesson sitting inside that one. “Banks lent 8% more” sounds like all banks doing more of a bad thing. The truth is a split: a chunk of banks moving one way, a smaller chunk moving harder the other way, and the average hiding the fight.

This is worth carrying past today. An average is a single number standing on top of a disagreement. When you read that “consumers are spending more” or “emissions fell,” ask what the spread underneath looks like — who went up, who went down, and which side was bigger. The headline number is real. It’s just not the whole shape, and the shape is where the story lives.

What breaks the trap, and why it’s hard

The escape from this kind of trap is never one player being braver. A single bank holding the line just hands business to the others and changes nothing. What breaks it is a rule that binds everyone at once — so the loan a good bank refuses can’t simply be written by a worse one.

That’s why the political shift in the briefing matters so much. Several banks had targets to cut lending; many quietly dropped them when the enforcement pressure eased. Take away the thing that bound them together, and each one slides back to the move that’s rational alone. The pledge wasn’t fake. It was just only ever as strong as the floor under all of them.

You’re already in the room

It’s easy to read this as a story about distant institutions and feel like a spectator above it. You’re not above it — you’re inside it, in more places than one.

The money those banks lend is partly yours: pension funds, savings accounts, the deposits that become loans. The fuel that gets drilled is why your electricity bill ran near $800 this summer, up more than 10%. And the heat that lending helps lock in is the heat a US study just warned could double hospital visits by 2040 — landing hardest on the places least built for it and the people who can’t afford to cool a room.

The trap isn’t out there. It runs through your bank, your bill, your summer. Seeing that doesn’t tell you what to do about it — and it shouldn’t make you feel clever for spotting the trick. It should make you hold your easy answers — “the banks should just stop,” “people should just choose better” — a little more loosely. Almost no one in this picture is choosing the whole. Each is choosing the one move that makes sense from a single seat, and the seats add up to something none of them would pick.

03 · Lab · your turn

The Loan Someone Else Will Write

Play one bank facing an oil loan and feel why restraint alone changes nothing — only a rule binding everyone does.