Why can't you just start a bank?
Once you know that banks create money when they lend, one thought arrives immediately and it arrives in almost everybody: fine, then I will start a bank. This page is about what actually happens when you try. It ends with a man who used to run research at a Federal Reserve bank, who designed a bank that could not lose your money, applied for the account it needed, waited more than six years, and was told no because it was too safe.
You can, and people do, but the door is narrow on purpose. In Britain you need permission from the regulators, a great deal of money up front, and roughly a year. Some of that exists to protect your deposits. But a bank designed to hold every deposit completely safe, lending none of it, has been refused outright, because the system is built to need banks that lend.
First, the ordinary door
Britain does not forbid new banks. It runs a whole unit to help people build them: the Bank of England's New Bank Start-up Unit, set up in 2016, which exists to walk applicants through the process. Anyone can read what it asks for. Here is the shape of it.
You submit an application and pay a fee of £56,300 before anyone has looked at your business properly. If your application is complete, the Bank of England aims to decide within four months, and has a legal deadline of six. If it is incomplete, the target stretches to ten months with a legal deadline of twelve. Two regulators have to agree; the Bank of England cannot authorise your bank on its own.
Most start-ups then go through a stage called mobilisation, where you get a restricted licence and build the actual bank: the systems, the staff, the controls, the money. During that stage your bank may not hold more than £50,000 of deposits in total. Not per customer. In total. Mobilisation is meant to last no longer than twelve months.
And that is only the paperwork. Behind it sits capital: real money you must put in and keep there, sized to the risks you take, so that when loans go wrong it is your money that disappears before anyone else's. Plus the people, the technology, the auditing, the reporting, and a plan that convinces two regulators you will still exist in five years.
It is worth saying plainly: a lot of that moat is there for you. A bank holds other people's money and creates money when it lends. Letting anyone do that with no capital, no controls and no scrutiny is how you get bank runs, and bank runs ruin ordinary people first. The barrier is real, and much of it is defensible.
Which is exactly why the next part is so interesting. Because the objection that stopped one particular bank had nothing to do with safety.
The bank that was too safe
Imagine a bank with one product. You give it money. It does not lend that money to anybody. It does not buy anything risky with it. It simply deposits every single pound at the central bank, where it sits as reserves, the safest form of money that exists, and it passes the interest back to you.
Nothing can go wrong with this bank in the ordinary way. It cannot make a bad loan, because it makes no loans. It cannot be caught out by a property crash. There is no maturity mismatch, no credit risk, no fire sale. A run on it is meaningless, because the money is already all there.
That bank was built. It was called TNB, The Narrow Bank, and its chief executive was James McAndrews, formerly head of research at the Federal Reserve Bank of New York. Not an outsider. A man who had spent his career inside the machine, who then designed the safest possible version of a bank and tried to open it.
To function, it needed one thing: an account at the Federal Reserve, called a master account, which is how a bank actually holds central bank money. It applied. And then it waited.
The answer, in the regulator's own words
More than six years after the application, the Federal Reserve Bank of New York said no. Its stated reasoning was that granting the account would "pose undue risk to the stability of the U.S. financial system and would adversely affect the Federal Reserve's ability to implement monetary policy."
Sit with that. The objection is not that the bank was reckless. It is not that customers might lose money. It is that a place where money is completely safe, paying something close to the central bank's own interest rate, would be too attractive. Money would leave ordinary banks and go and sit there, especially in a crisis, when everybody wants safety at once. Other reasons were given too, including that it had no deposit insurance and none of the normal banking connections. But the headline objection is about the system, not the customer.
Go back to the fact this whole site is built on. New money is created when banks lend. It is not printed and handed out; it appears in your account as a bank's promise, at the moment a loan is written.
Now put the two things together. If money moved out of lending banks and into a perfectly safe warehouse, the machine that makes new money would have less to work with. The safety of a narrow bank and the money-creating power of an ordinary bank are the same question asked twice. You cannot have a system built on bank lending and also make it easy to opt out of bank lending.
That is not a moral judgement. It is a description of how the thing is wired. But it does mean the honest answer to "why can't I just start a bank?" is not only "because it is hard and expensive". It is also: because a bank that refuses to create money is not the kind of bank the system is designed to have.
What this does and does not prove
It does not prove that regulators are corrupt or that the refusal was wrong. Serious economists argued both sides in public, and some think a narrow bank really would be destabilising in a panic, drawing money out of the banking system precisely when it can least be spared. That is an argument worth taking seriously.
It does prove that the barrier to starting a bank is not only about protecting depositors. At least once, a bank was refused specifically because it was too safe and would work too well. When you next hear that banking is fiercely competitive and anyone can enter, you now know one very well-qualified man who spent six years finding out otherwise, and can read the letter.
- Read the Bank of England's New Bank Start-up Unit pages. They are written for applicants, not for the public, which is why they are so revealing. The fee, the timescales and the £50,000 mobilisation cap are all stated there.
- Look up how many new UK banks were authorised last year. The number is public. Compare it to how many high street names you can list.
- Check what protects your own money. UK deposits are covered up to £120,000 per person per authorised firm. Worth knowing which firms share a licence, because the limit is per firm, not per brand.
- Search "TNB USA master account". The application, the refusal and the arguments on both sides are all in public, including the reasoning quoted on this page.
The Bank of England's New Bank Start-up Unit, on bankofengland.co.uk, for the £56,300 application fee, the four-month target and six-month legal deadline for complete applications, the ten and twelve month versions for incomplete ones, the mobilisation stage, and the rule that a bank in mobilisation may hold no more than £50,000 of deposits in total.
The Federal Reserve Bank of New York's refusal of TNB's master account application, reported in the banking trade press in February 2024, quoting the letter: granting it would "pose undue risk to the stability of the U.S. financial system and would adversely affect the Federal Reserve's ability to implement monetary policy". Search for TNB USA and master account and read several accounts, including the ones that think the Fed was right.
The Bank of England's own 2014 bulletin on money creation, for the claim underneath all of this: that most new money is created by commercial banks making loans. It is the document that makes this page make sense.
Every figure above was checked on the day this page was written. If one has changed, tell us and we will change it.
Questions people ask
Can anyone actually start a bank in the UK?
Legally yes, and new banks do get authorised. Practically you need two regulators' permission, a £56,300 application fee before anyone examines your plan properly, substantial capital, and roughly a year, including a mobilisation stage during which the bank may hold no more than £50,000 of deposits in total.
What is a narrow bank?
A bank that holds all its deposits as central bank reserves and lends none of them. It cannot make a bad loan because it makes no loans. It also creates no new money, which is the reason it is controversial.
Why was TNB refused?
The Federal Reserve Bank of New York said granting it an account would "pose undue risk to the stability of the U.S. financial system and would adversely affect the Federal Reserve's ability to implement monetary policy". Other reasons included the absence of deposit insurance and of normal banking links. The refusal came more than six years after the application.
Is my money safe in a normal bank then?
Your deposits are protected up to £120,000 per person per authorised firm if the bank fails. That protection is real and it is why most people never have to think about any of this. It is also a promise standing in for the fact that the money itself has largely been lent out. Both things are true.
Does this mean the regulators are protecting the big banks?
Some people argue exactly that, and some argue the opposite: that a completely safe alternative would drain deposits from ordinary banks in a panic and make a crisis worse. Both arguments are public and both are serious. This page gives you the documents and the question; the conclusion is yours.
Read next
- Rabbit holeHow do banks create money? (Yes, out of nothing.)
- Rabbit holeWho actually owns the banks?
- Rabbit holeJekyll Island: the secret meeting that wrote the Federal Reserve
Last checked 25 August 2026.