The Old Guard's Dilemma

The quiet war the headlines miss

Follow the money far enough and you reach a fight quieter than the one on the news – between the institutions that profit from today's money and an asset designed to need them less.

The public version is theatre. The most powerful banker in America – JPMorgan's Jamie Dimon – spent a decade calling Bitcoin a "fraud," "worthless," a "pet rock" whose only real use is crime – and then, in 2025, had his bank begin letting clients buy it, with the line: "I don't think you should smoke, but I defend your right to smoke."¹ The sneer is tempting evidence, but it proves less than it seems. People mock plenty of things that are merely dumb-but-popular and never lift a finger to sell them – beanie babies, meme stocks, the lottery. Mockery alone tells you nothing.

The evidence that matters is not the sneer; it is the sale. The same firms that called it a fraud now custody it, list it, and put it on client statements – they built the thing they ridiculed. Watch what an institution builds, not what its figureheads say, because building costs money, lawyers, and reputation, and no one pays that price for a passing fad they could simply wait out. A pet rock you ignore. You do not staff a desk for it.

A pet rock you ignore.
You do not staff a desk for it.

The loudest skeptic in banking, in his own words¹ – and then his own product line.

There is an innocent reading, and it deserves a fair hearing: maybe the banks are simply selling a product clients want, the way they would sell tulip futures if the orders came in, with no conviction either way. Demand exists; you serve it; the sneer is just an honest opinion of a silly fashion. That story is plausible until you ask what, specifically, the banks are giving up to serve that demand – because a bank does not custody a meme stock, lobby over a beanie baby, or rebuild its settlement plumbing for the lottery. The intensity of the response is the tell. You only fight this hard, this long, and then capitulate this completely, over something that threatens a rent – a structural stream of income built into the old system. Client demand explains a desk. It does not explain a decade of organised resistance followed by a scramble to own the rails. For that you need a reason the asset costs them money, and the next section is that reason.

What you cannot see is the rest of it – the lobbying, the regulatory friction, the quiet weight a trillion-dollar incumbency can put on a thing it cannot own. None of that arrives in a press release, so treat its specifics as informed speculation rather than reportage; the shape, though, is plain in the public record: scorn on the stage, capitulation in the product line.

What they actually stand to lose

To see why an industry would resist, look at what today's money pays it – and what a bearer asset quietly takes away. There are three rents, and they are the real subject of this essay; the Dimon theatre is only the door into it. Name them precisely, because they are the premises the whole argument runs on: if a bearer asset truly drains all three, the behaviour we observe stops being a puzzle and becomes the predicted result.

A bank's deepest business is intermediation: standing between you and your money and charging, in a hundred quiet ways, for the privilege. The fees you can see on a statement are the small part. The large part is structural – income that flows not from a service rendered but from a position held, the position of being unavoidable. A bearer asset attacks that position directly. Three rents are at stake.

The first is the debasement premium. New money does not reach everyone at once. The institutions nearest its creation – central banks, then the commercial banks and large holders downstream of them – receive it first, at full value, and spend it before the prices it inflates have caught up. Everyone further from the spigot receives the same currency later, already diluted. This is the oldest privilege in finance: the gift of being early to freshly printed money, paid for invisibly by everyone who is late. A supply that cannot be expanded – twenty-one million, no committee, no exception – simply abolishes that gift. There is no "first" to be near when nothing new is ever made. For the institutions whose franchise is partly built on proximity to the printer, that is not a fee they lose; it is a form of income that ceases to exist.

The second is the deposit franchise, and it is the largest. The old textbook picture – the bank takes your deposit and lends "it" out many times over – is not quite how it works, and the truth is more revealing. A bank does not need your money sitting there first; when it makes a loan it simply writes a new deposit into existence on its own books, constrained not by some pile of cash on hand but by its capital, its regulators, and how many creditworthy borrowers want to borrow. What your deposit actually buys the bank is something subtler and more valuable: a vast, cheap, sticky pool of funding it pays you almost nothing for, against which it runs the whole lending machine and earns the spread. And it stays only because moving it is friction, because where else would it go. A bearer asset you custody yourself removes the friction and answers the question. It is money that can simply walk out of the building and keep working for you instead of for the bank. Multiply that across a balance sheet and you are not trimming a fee; you are draining the cheap funding base the entire franchise is priced on.

The third is the toll roads – the spread on moving and storing money. Wires, settlement, foreign exchange, correspondent banking, custody: every time value crosses a border or changes hands, an institution sits at the crossing and takes a cut, and because the roads are few and privately held, the toll is whatever the road owners agree it should be. A network that settles itself, for anyone, anywhere, at any hour, without asking permission, does not abolish the road – it just ends the monopoly on it, and a monopoly toll collapses to a competitive one the moment a second road opens. The fee survives; the markup does not.

Their business is being the middleman.
This is money that needs no middle.

There is a fourth, softer asset at risk, and it is the one that ties this essay to the rest of the argument: trust as a chokepoint. When money lives only on ledgers the banks keep, the banks decide what clears and what is frozen – the same gatekeeping power that lets reserves be seized and accounts be cut off, examined elsewhere on this clock. That power has, more than once, been abused. Consider what LIBOR was: a benchmark that set the price of money on trillions in loans and contracts worldwide, and a cartel of the largest banks was caught rigging it for years – manipulating the very number that priced the world's debt.³ It worked precisely because a handful of institutions sat astride the plumbing and quoted the figure to themselves. That is the whole case for a money no committee sets, compressed into a scandal: the institutions that price money have been caught lying about the price. A ledger that verifies itself does not ask the banks what time it is, and cannot be quietly told to read the clock wrong.

None of which means the banks vanish. Most people will always want someone to hold the keys, undo the mistake, answer the phone, and that service is worth paying for. What erodes is not banking but the monopoly on it, and the fat, invisible rents that came from being the only road money could travel. A smaller, honester toll is still a business; it is just a far less lucrative one. That gap – between the rent they earn now and the fee they could earn after – is the whole financial stake of the fight.

Why the cannier ones are already joining

Here is the spine of the whole essay, and it needs no peek inside any banker's head to hold. Set intent aside entirely. Whatever any executive privately believes, the three rents are eroded by a single property of the asset – it is a bearer instrument that any holder can self-custody and settle with finality, needing no institution to clear it. That property does not negotiate. And it forces a dilemma with exactly two branches, no third, because an incumbent confronted with a rent under technological attack can do only one of two things: cede the rent, or re-intermediate it. Both branches compress it.

Trace each rent down both branches and watch the conclusion fall out structurally. Cede means do nothing – let clients hold their own coins, settle their own payments, sit outside the franchise. The rent is simply gone: no spread on money you never touch, no funding base in coins that left the building, no toll on a road you do not run. Re-intermediate means build the rails yourself – custody the asset, wrap it in a fund, run a stablecoin, charge for access. This recaptures a fee, but a different one: the asset on those rails still has, sitting beside it, a free exit the bank cannot remove – self-custody. And a fee charged on a service the customer can walk away from for nothing is a competitive fee, not a monopoly rent. It collapses to whatever a second provider, or the footpath, will undercut it to. So the cede branch zeroes the rent and the re-intermediate branch caps it at the competitive floor. There is no branch on which the old rent survives – which is why this is a dilemma and not a choice. The conclusion does not depend on what the banks want; it depends on what a bearer asset with a free exit does to any toll laid on top of it.

Cede the rent, or re-intermediate it.
There is no third door – and both doors compress it.

That structural fork is exactly what the industry is visibly splitting along – not because each bank reasoned it through, but because there were only ever the two branches to fall down:

The participants have simply read the board. If bitcoin is going to be owned, it will be owned through someone's product – so better it be yours, earning a custody or trading fee, than a rival's. Morgan Stanley chose that path early: its own low-fee spot bitcoin fund, a few-percent crypto allocation pushed out through sixteen thousand advisors, trading wired into E*Trade for millions of ordinary accounts.² It is the innovator's dilemma in real time – the rational incumbent move is to cannibalise your own franchise before an outsider does it for you.

If the river is going to flow,
be the riverbank.

There is an irony the maximalists savour: an ETF is bitcoin with the bank put back in. Buy the fund and you have re-hired the very intermediary the asset was built to remove – paper claims on coins a custodian holds for you. That is exactly why the banks can make their peace with the ETF: it lets them sell the new asset on the old terms, fee intact. The fee economy migrates to them; the sovereignty does not. The institutions are not leading the public toward bitcoin; they are being pulled, building the rails for a thing many of their own leaders still claim to disdain.

Now the strongest reason a sincere, honest bank stays out – stated at full strength, because it is real and most of the early caution was genuinely this. A regulated institution that holds an asset for you takes on liabilities the asset itself does not carry. Bitcoin is volatile enough to detonate a quarterly result and a client relationship in the same week. Its retail history is thick with hacks, frauds, and exchanges that vanished overnight, and a custodian is on the hook when keys are lost or stolen, with no chargeback and no recourse. Capital rules have, for years, treated crypto holdings as near-radioactive on a balance sheet, demanding punishing reserves against them. And a fiduciary's first duty is not to chase the new thing but to avoid ruining the client – so moving slowly on an asset this raw is not cowardice, it is the job. A great deal of the early "resistance" was exactly this, and it was correct. Not every objection is a defence of rent.

So how do you tell sincere prudence apart from rent dressed as prudence? You watch what changes when the prudent reasons fall away – and you hold the two innocent explanations to the same test the rent explanation must pass. Lay all three side by side against the one thing we can actually observe, which is the sequence: a decade of organised disdain, then a capitulation timed precisely to when the rents came under threat rather than to any change of heart, then a race for the toll booths. "Merely serving demand" fails on the front of that sequence – it cannot explain the years of resistance that ran ahead of the demand, because you do not lobby and slow-walk a product your clients have not yet asked for; a desk that only follows orders has nothing to resist. "Merely prudence" fails on the back of it – as custody matured, regulation clarified, and the capital treatment eased, a sincerely fiduciary firm would resolve into a shrug, fine, we'll offer it, modestly, like any other instrument; it would not scramble for funds, custody, trading rails, and advisor allocations the instant the danger passed. Prudence explains the waiting and predicts a calm exit from it. It does not explain the hunger that erupted once the waiting ended. Only the rent explanation accounts for both ends at once: you fight a thing that threatens a structural income stream, you fight it for as long as the franchise holds, and then you race to own the replacement so the new fee at least accrues to you. Interest hides inside conviction, and the way you separate them is to watch which motive survives the moment the excuses expire. Run that test and the verdict is not a coin-toss held level with its rivals: the rent-defence reading is the only one of the three that fits the whole sequence rather than half of it. That is an inference, not a confession – the banks will never say it – but it is the conclusion the evidence actually forces, and the honest uncertainty lies not in why they behaved this way but in how far the erosion ultimately runs.

Here is the honest counter to this essay's own thesis, and it is not weak: the participation could be the system winning, not surrendering. If almost everyone ends up holding bitcoin through a bank's fund, the incumbents have quietly re-intermediated the one asset built to remove them – collecting the fee, controlling the access, holding the keys, even able to freeze or lend out the paper claim. On that reading the dilemma resolves in their favour: they defang the threat by absorbing it, and the revolution ends as another line on a custodian's balance sheet.

But the re-intermediation cannot fully close, and this is where the argument tilts. Every prior monetary order the banks captured was captured because there was no exit – you could not opt out of holding dollars in some bank somewhere; the asset and the custodian were the same thing. Bitcoin breaks that for the first time: the asset exists independently of any institution that holds it. The one thing the banks cannot capture is self-custody – twelve words written down, a thing no fund can offer and no bank can charge for, the person who holds their own keys has simply left the system. That exit is permanent and it is free, which means the bank's product can only ever be a convenience layered on top of an asset that does not need it, never the asset itself. The toll road has a footpath beside it that no one owns. So the ETF can re-intermediate the lazy majority, yes – but it sets the ceiling on the rent, because the moment a custodian's toll or its freezing power grows intolerable, the keys are right there. The middleman returns, but on a leash this time – held by anyone willing to learn twelve words. How far the majority walks down that footpath is the genuinely unsettled question – the one honest unknown in this essay. Which way the floor sits is not unsettled: it rests on self-custody, and that floor does not move.

And here is the turn most people miss, the one that makes the banks' surrender stranger than it looks. When JPMorgan launches a fund, when an institution that spent a decade calling bitcoin a fraud quietly puts it on the menu, it is not merely conceding a business line – it is lending the asset the one thing the asset could never manufacture for itself: institutional legitimacy. A thing the gatekeepers custody and sell is, by that very act, declared real and bankable. So the capitulation and the profit are not two events but one motion: the same move that lets the bank collect the new toll is the move that ratifies the asset to every cautious trustee, pension board, and regulator still waiting for a respectable signal. They cannot monetise it without first blessing it – and the blessing is worth more to bitcoin than the toll is worth to them. The incumbent's defeat and the incumbent's fee arrive in the same gesture, which is why this is the rare battle the loser helps win.

They cannot sell it without first blessing it
– and the blessing outlasts the fee.