Hacker News Reader

Comment by stephen_g | Hacker News Reader

This article has a surprising misunderstanding of the way banks work for a publication called “The Economist”. It is unintuitive, but banks do not and can not lend out deposits. Banks lever their capital (that is, paid up shares and retained profits) to lend, and that lending creates new deposits. Existing deposits and debt funding (money markets etc.) are useful as liquidity so they can settle interbank transfers, but they aren’t allowed to lend out of them, both out of regulations like the Basel framework, but also because it would violate accountancy rules (deposits are liabilities, and you can’t back a liability with a liability. In reality they create the loan (asset) with a corresponding liability (the deposit).

The Bank of England (basically the UK federal bank) has published a good article on this - https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m...

View on HN

Banks absolutely do lend out deposits. An average bank will take a $100K in deposits, have $10K in equity, and then lend out $110K in a mortgage.

> Banks lever their capital

And with what do the banks lever up their capital? (What is the indirect object of that sentence?) The answer, of course, is deposits.

Take an example of a bank local to me: http://ir.cambridgetrust.com/AsReported/Index

They have about a $1.370B in deposits. Where do they put it? Look in the assets section and you'll see it's all commercial and residential mortgages.

That’s a common misconception. What you have described is called the money multiplier model, and is absolutely incorrect. That’s exactly what the Bank of England paper is saying.

An additional point of reference is the Basel rules, which are pretty clear on this, if you care enough to read a few dozen pages of banking regulation.

> “And with what do the banks lever up their capital? (What is the indirect object of that sentence?) The answer, of course, is deposits.”

Incorrect. It’a a little more abstract than that. Bank credit doesn’t exist as anything but numbers in a database. Deposits are IOUs from the bank, which are increased when the bank lends, and mortgages are IOUs from people to the bank. If everything just stayed within the single bank, they wouldn’t need any existing deposits at all, or any reserves! What you need the reserves for (one source of which is incoming deposits from other banks, or cash deposits, or cheques from the Government) is interbank settlement. But the bank needs only a small fraction of the actual volume of transactions, because they are generally going both ways, and transfers are only a small fraction of the actual amount of bank credit that exists. That’s how the bank maintains solvency.

You don't need reserves for interbank settlement either. That's actually a centralised optimisation to reduce collateral requirements.

At root if you transfer your deposit to another bank that target bank becomes the depositor in the source bank. Or the transfer can't happen.

Banks then shuffle those deposits they have take on around until they are happy with the risk. That's interbank clearing.

Then how do the banks profit from deposits, if they can't use them for lending?

They don't. They might get some interest on the cash you gave them when you deposited the money, or for a transfer the reserves that they got from settling deposits from the central bank, or from lending those reserves to another bank. Or they might convert those assets into bonds etc.

But the 'bank deposit' itself is just an entry on the liability side of the bank's balance sheet, they can't profit from that. The bank might pay interest on it to the depositor because they want to encourage them to deposit more, because deposits coming in give the bank cheap liquidity, so stop them having to borrow reserves from other banks etc. to maintain liquidity.

Why would a bank take deposits at all then?

I'm not saying you're wrong. It just seems like something is missing from the picture in your explanation.

When I opened my bank account, I deposited say £5000 cash. What does the bank do with this money?

I was under the impression that the bank doesn't actually keep this £5000 in cash, only £500 maybe, and uses the rest to lend money, but you seem to be saying that they can't actually use my deposit.

"When I opened my bank account, I deposited say £5000 cash. What does the bank do with this money?"

It burns it. Probably quite literally.

The cash represents a transfer from the central bank into the commercial bank's deposit account at the central bank - for which it receives an interest rate.

Cash is just a receipt for deposits held at the central bank.

Deposits are a relatively cheap way of attracting reserves. When you added £5000 from say an account at Santander, Santander lost a liability (the deposit) and simultaneously transferred some reserves to your current bank.

Banks can also attract reserves by borrowing in money markets (cheap but short term) or issuing long term debt (stable but expensive), or by persuading customers to open accounts and keep them in positive balance. In practice it will use a mix of all three, attempting to match the maturity profile of its liabilities.

The central bank will also create reserves and lend them to your bank if your bank can't raise them elsewhere for some reason. But borrowing from the central bank can spook investors, and it's also expensive.

A bank that doesn't have reserves can't settle its debts to other banks (like the debt you incurred by moving your £5000) and so can't function in the banking system. It has to get reserves from somewhere, and persuading consumers to park deposits in an account is one useful way to do so.

I am not sure where you saw that banks are not using deposits to fund loans. For instance if you look at lloyd's balance sheet [1]:

  loans                            482,752
  cash and afs (liquidity pool)    100,619
  trading assets                   163,878
  other                             64,860
  total assets                     812,109
                     
  Deposits                         447,928
  trading liabilities               50,877
  wholesale funding & capital      140,828
  Insurance liabilities            103,413
  other                             69,063
  total liabilities & equity       812,109

I believe the trading assets, trading liabilities and insurance liabilities must have to do with their insurance business and sort of net each others. What is left is basically 450bn of deposits and 140bn of capital and wholesale funding, funding 480bn of loans and 100bn of liquidity.

Now cash is fungible, and you cannot point to a particular asset and say that a particular liability is funding this asset. But overall, deposits are funding loans.

[1] https://www.lloydsbankinggroup.com/globalassets/documents/in...

Only 1) the Basel framework of bank regulation which has been internationally adopted, and 2) explanations from the Bank of England (linked above) which categorically states that this is actually how banks work. The Deutsche Bundesbank has also published something to this effect, and the Governer if the Reserve Bank of Australia talked about it in a speech the other day [1].

1. https://www.rba.gov.au/speeches/2018/sp-ag-2018-09-19.html

I don’t want to get into some ad hominem attacks but I don’t think Basel 3 says that. Anyway, this is simple arithmetic, Lloyds could not have such a large loan book if it wasn’t funded by deposits. There are simply just not enough wholesale funding and capital liabilities. And the liquidity requirements are only a fraction of the deposits, particularly for retail deposits which are assumed to be sticky.

It absolutely does say that. The whole of part 1, "Minimum capital requirements and buffers" is 100% based on capital, and deposits don't factor in to any of the definitions of Tier 1 capital. The leverage ratios also factor entirely from capital that doesn't include deposits.

100% is the conversion of a loan balance to leverage exposure. The capital requirements vary between banks and regulator, but they are typicaly 3-5% of the leverage exposure. A leverage ratio requirement of 100% would make lending uneconomical. To achieve a return on equity of 10%+, you would have to charge over 10% on your loans.

My understanding is that deposits are really just a manifestation of a promise that a trusted party holds, so that the party making the deposit can make more promises to other parties. So, in banking a promise is often just as valid as holding a deposit. They aren't really lending you value, they are lending you verification of value.

They are unsecured liabilities. But like all liabilities the bank received an asset when it issued it (cash or else), and like all liabilities the bank must have sufficient assets to pay it back on demand. The bank may not have the necessary liquidity to pay back all deposits at once (some of these assets are 30y mortgages) but it has an asset (otherwise is not solvant).

When lloyds creates another loan it automatically creates the matching deposit. You get both a mortgage debt and a credit asset when you take out a loan. You then transfer that credit asset to somebody else to pay them. But it is the loan that creates the credit asset in the first place. Deposits have no quantity control function on lending. It's only ever a pricing issue.

When you lend to someone, he is free to transfer the money immediatly to whomever, who may or may not be a customer of that bank. The loan must have been funded for the bank to be able to afford to transfer the deposit away. Now you may get some deposit back that same day but the funding decision has to be made before.

Anyway, neither funding or pricing determines the amount of lending these days, capital ratios are the limiting constraint.

The bank needs liquidity to settle inter-bank transfers after they've lent, some of that is deposits but they can also just borrow reserves from other banks if they need to do that.

The lending itself created new bank credit, it was never "funded". If the bank has capital adequacy, they make the loan.

Ah, but when they transfer that money, it'll be to some customer of some bank whose loan customers will also be transferring money in the other direction. This means, in practice, that the main constraint on a bank's ability to lend money is how willing all the other banks are to lend money at the time. The Bank of England paper linked earlier in the thread explains why this has undesirable economic implications that are missed by the incorrect money multiplier model.

When the deposit is transferred away essentially the target bank becomes the depositor in the source bank.

That's effectively what 'interbank lending' is. The central bank system then centralises that peer-to-peer processes.

Deposits are swapped for capital in the same way.

> You then transfer that credit asset to somebody else to pay them.

So the matching deposit disappears from the liabilities side of Lloyd’s balance sheet while the loan remains in the assets side? That leaves the balance sheet unbalanced, doesn’t it?

The deposit may disappear from lloyds' liabilities, but it needs to transfer an equivalent amount from its reserve account to (say) Santander's reserve account. So Lloyd's loses both liabilities and assets when someone moves deposits to another bank.

Of course, my point is that neilwilson’s statement (emphasis mine): “When lloyds creates another loan it automatically creates the matching deposit. (..) Deposits have no quantity control function on lending.” is at best misleading.

A bank with an empty balance sheet cannot give a loan.

Lending money is transforming one asset (reserves) into another (loan). The fact that the loan is created as a new deposit in the name of the borrower, keeping the reserves constant, muddles the issue a bit but the simplest description of the process is something like:

1) D makes a $1mn cash deposit at the bank. The bank has liabilities: $1mn owed to D and assets: $1mn in cash.

2) The bank lends $1mn in cash to B. The bank has liabilities: $1mn owed to D and assets: $1mn lent to B.

We can add an intermediate step where the loan is still a deposit at the bank (liabilities: $1mn owed to D. + $1mn owed to B., assets: $1mn in cash + $1mn lent to B.) but it is just a distraction. Not many people borrows money to keep it in a deposit at the same bank.

That's not the case.

A bank with an empty balance sheet can give a loan. It is simply Dr Loans, Cr Deposits. That's it.

Nobody ever makes cash deposits into a bank. The cash is just burned when the bank receives it (if it is worn enough).

A cash deposit to a bank is just a deposit transfer from the central bank. Cash is just a receipt for entries already at the central bank. You are starting half way down the process.

To take your example 1) The balance sheet is $1mn Fed deposit asset, $1mn Credit to D. 2) The Balance sheet is $1mn Fed deposit asset + $1mn Loan Asset to B, $1mn Credit to B, $1mn Credit to D. 3) If the cash is withdrawn then the Fed deposit asset and the Credit to B is marked down. The cash will be new cash issued as a receipt for the deposit at the Fed.

You can do exactly the same without the deposit of D first. At that point the Fed asset would be negative and become the lender with the loan posted as collateral.

> A bank with an empty balance sheet can give a loan. It is simply Dr Loans, Cr Deposits. That's it.

A banks needs capital, an empty balance sheet doesn’t cut it.

> Nobody ever makes cash deposits into a bank.

Yes, they do. If you pay $5 for a capuccino or whatever what do you think they are going to do with your $5 bill at the end of the day?

> The cash is just burned when the bank receives it (if it is worn enough).

So they never receive cash because nobody ever deposits cash. And at the same time they do receive cash and they just burn it. You have convinced me, there is something seriously wrong with the banking system.

I'm not really sure what you mean. The Bank of England article you linked says:

> that would leave the buyer’s bank with fewer reserves and more loans relative to its deposits than before. [...] > By attracting new deposits, the bank can increase its lending without running down its reserves, as shown in the third row of Figure 2.

So actually the banks need deposits to hand out loans. What you are saying is that the deposits get a new label before handing them out as loans?

Banks need the liquidity from various sources (including deposits) because they lend, not to lend.

On the bank's balance sheet, they cannot take deposits and lend them out.

Don't banks take deposits because they want to lend?

Sure, it's fairly cheap liquidity, the more the better. But no bank ever stops lending because they don't "have enough" deposits, because there is nothing of the mechanics of lending that relies on having existing deposits.

They just need to be 1) solvent and 2) within their capital adequacy requirements and leverage ratios.

Of course they do. That's what fractional reserve banking is all about. Banks borrow money (e.g. from depositors, on the money markets, from other banks) and lend out the money they raise while keeping a "reserve".

A typical bank's balance sheet will contain deposits on one side and loans on the other. Without deposits, or other sources of borrowing, a bank cannot lend money.

Again, that’s just a popular misconception. The document I linked to before from the Bank of England [1] is all about how what you describe, while commonly believed, is not how modern, real world banks work.

1. https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m...

The document you linked does not contradict my claim. Your document describes the role of banks in the M2 money supply NOT how individual banks work.

I can assure you that what I describe IS how real world banks work. You can look at the balance sheet of any retail bank and it will spell this out clearly. I've worked in retail banks and have some hands one grasp of their day to day operations.

Your confusion is a common one - you mistake the process that results in the creation of M2 money with the day-to-day operations of retail banking.

Modern retail banking is a very simple business - they borrow funds (from depositors, other banks, sometimes central banks, commercial loans, etc.) and then lend out the SAME funds while keeping some back in liquid assets (cash, deposits in central banks, short term government notes, etc.) in reserve. The trick is to charge more interest on what you lend out than you pay on what you borrow while managing the risk and cash flows involved when you have a mixture of terms (expiries) of borrowings and lendings.

I think I've given a reasonable summary in my posts of how banks work. I've studied the balance sheets of retail banks and have been exposed to their day to day operations so please don't simply disagree with me and point me to a document on M2 money creation - if you think I've stated something untrue, then highlight it and describe where I'm wrong.

Banks are not merely intermediaries between savers and borrowers. Loans create money, but only loans that are not used to repay other loans result in an increase to the money supply. The following text comes from an ING Bank research note, quoted by The Economist [1]:

"Banks do not view the creation of money as an objective itself. It is a by-product of the banking sector’s business operations. However, it is of great economic and social relevance.

Not every loan ultimately results in new money. The majority of new lending is used to redeem existing loans. Money is only created to the extent the gross lending exceeds the value of the existing loans being redeemed."

That note refers to the "great economic and social relevance" of these banking operations. Here's Professor Richard Werner talking about this at length. [2]

Here's Perry Mehrling (who teaches Coursera's Economics of Money and Banking) weighing in [3]. He explains that it's a nuanced issue but clearly agrees that the "credit creation view" is important and quotes a Group of 30 report:

“In a barter economy, there can rarely be investment without prior saving. However, in a world where a private bank’s liabilities are widely accepted as a medium of exchange, banks can and do create both credit and money. They do this by making loans, or purchasing some other asset, and simply writing up both sides of their balance sheet.”

[1] https://www.economist.com/buttonwoods-notebook/2014/06/11/wh...

[2] https://www.youtube.com/watch?v=N-FDdHj7rPk

[3] http://www.perrymehrling.com/2016/01/great-and-mighty-things...

Again banks' role in the creation of M2 money does not reflect how individual retail banks operate.

Just because you have a model that seems to describes the behavior of flock of birds doesn't mean that the model is useful when discussing the anatomy and wing-aerodynamics of individual birds.

You've posted a bunch of links describing the various forms of money and the role of banks in modern (post 18th century) M2 money creation. I know all of this material but it's not relevant to the operations of individual retail banks.

OK, so if each individual retail bank only lends out pre-existing, deposited funds, where does the increase in the money supply come from? Bear in mind that money created as a by-product of bank lending makes up the vast majority of the total money supply in various modern economies (97% in the UK [1]).

[1] https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...

The increase occurs because of the different definitions of money. If you define money as the sum of amounts available in demand account at banks (with is roughly what M2 is), then lending from a bank increases this number as the borrowed money ends up in the account of the borrower or in the accounts of the people the borrower spends the money with.

This fact means that retail banks "create money out of nothing" only when money is defined as the sum across all banks of the amounts available to customers on demand. No individual bank "creates money out of nothing", individually they borrow money and lend a fraction of it out; they have balance sheets which balance - i.e. ignoring shareholder equity, for every asset (money owed to them by a borrower), there must be a liability (money they owe to lenders including depositors). The M2 money supply counts the latter but does not deduct the former from the sum; as a result, the M2 money supply has a direct relationship with the total size of retail banking balance sheets.

M2 is an economic statistic and it's behavior tells you nothing about how individual banks operate.

OK, how does that gel with a report from S&P [1] that says:

“Banks lend by simultaneously creating a loan asset and a deposit liability on their balance sheet. That is why it is called credit “creation” – credit is created literally out of thin air (or with the stroke of a keyboard)”?

Also Prof Werner's analysis, he reaches the same conclusion.

[1] http://positivemoney.org/2013/08/repeat-after-me-banks-can-n...

That statement is misleading or reflects a misunderstanding on the part of the author.

"Credit" is an accounting concept. Balance sheets are accounting artifacts. The way the bank records it's transactions is a model, not reality. For example, lots of intangibles can have accounting entries created for them for various purposes, but these don't create reality.

Btw, a big problem when discussing bank operations are that the terms "debit" and "credit" can be confusing as their roles are relative to the bank not the customer. Even "asset" and "liability" can be confusing. An Irish media finance and economics professor, Brian Lucey, mixed them up and suggested that a troubled bank could be fixed if it sold its deposits which is impossible - deposits are a liability for banks, not an asset.

Reality is much more simple. Bank lending starts when a customer asks for a loan. Assuming the loan application meets all risk requirements, etc. AND sufficient funds (reserves) are available on the bank's balance sheet, the loan can be issued and the funds transfered to the customer. Reserves are reduced as part of this process but an asset is created for the bank (the loan) so the balance sheet remains the same size.

>> the funds transfered to the customer

Well, that's the crux of it right there. You are disagreeing with several authoritative sources who say that no such transfer happens, and in general the loaned funds are created out of thin air.

Standard & Poor's: “Banks lend by simultaneously creating a loan asset and a deposit liability on their balance sheet. That is why it is called credit “creation” – credit is created literally out of thin air (or with the stroke of a keyboard)”

Group of 30: "... banks can and do create both credit and money. They do this by making loans, or purchasing some other asset, and simply writing up both sides of their balance sheet."

Richard Werner: "...each individual bank creates credit and money newly when granting a bank loan."

Banks are subject to constraints. They need to retain enough capital to absorb losses on their loan book, and they need to retain enough reserves to cover withdrawals and clearing requirements. They are free to create money "out of thin air" insofar as they meet regulatory requirements associated with those constraints. [1]

You wrote: "I've worked in retail banks and have some hands one (sic) grasp of their day to day operations."

I suggest that your experience does not encompass the whole sector, and that may be the source of your confusion.

[1] https://www.goodreads.com/book/show/13144133-where-does-mone...

Maybe it's just me but this really seems like quibbling over semantics. Cash is fungible: banks need cash in order to lend it out and they get cash from deposits. The rest is just bookkeeping.

There's clearly not enough in deposit to sustain this model. In the US, a large majority of people don't have enough saved to last 3 months.

banks need deposits to keep their leverage ratio in check.

and due to overleveraging one unit of deposit ends up as multiple units of loans.

and sure, deposits are not the main factor for lending, risk is. but usually taking deposits is cheaper than taking out a loan from an institution and lending that out. hence banks want/need deposits.

Ugh. It is a bit like saying that people don't earn money to buy things, but earn money because they buy things! By this line of reasoning when I go to a store, I am totally unconstrained by how much I earn - I just take my credit card, swipe it and buy whatever I want. Now, this strategy would be unsustainable in the long run if not backed by sufficient inflow of cash from my employer. So I have to work every day. But these are totally separate processes!

A lot of bookkeeping gets done, assets and liabilities change hands, but in the end it can still be described simply as "people earn money and buy things with it."

> So actually the banks need deposits to hand out loans. What you are saying is that the deposits get a new label before handing them out as loans?

Well, no. What the BoE paper is saying is this: Suppose there are two banks, A and B.

If Bank A went on a lending spree without attracting new deposits (or otherwise bringing in new reserves), many of the deposits it created would end up with bank B, because bank A has no control over how its deposits are used. Bank A would have to transfer reserves to bank B to settle the transactions. Eventually bank A would run low on reserves and would need to raise more (by issuing debt, or by making its deposit rate more attractive).

But if bank B was going on a similar lending spree at the same time, the two banks wouldn't run down each others's reserves. If the two banks act in tandem (e.g. by getting caught up in speculative mortgage lending), they can create plenty of money without needing to attract new deposits.

The biggest regulatory hurdle that stops this happening is that there's a regulatory constraint on the leverage ratio (assets - liabilities) / assets.

> In reality they create the loan (asset) with a corresponding liability (the deposit).

Isn't that just a book-keeping technicality? Fundamentally, the bank is still using deposits to leverage a loan. Or am I misunderstanding your point?

So basically I deposit $1000.

Someone else goes to the bank and says ‘I want to borrow $1000’

Great the bank says, approved. And they set up a bank account for you that says you have $1000 in it.

Now they’ve got $1000 in reserves and 2 accounts that claim to have $1000 in it.

Now when you withdraw $500 where does that money come from? I can assure you it’s not from the banker’s wallet. That’s coming out of my deposit. Not from my account because my account still says they have $1000 on hold for me, it comes from the actual cash I deposited. (Which of course isn’t actually cash in most cases)

None of that matters in practice because unless everyone tries to withdraw all the money at once there’s plenty to cover all the withdrawals, but they are definitely using money from deposits to cover withdrawals.

You can't deposit a $1000 unless either a bank or the state created it first. Those greenbacks are just receipts for a deposit at the Fed created by prior government spending.

I think part of the problem to understanding is that you’re assuming that banks are places that store money. They aren’t really, they are places that hold various types of assets and manage IOUs between themselves and other parties (other Banks, the Government/Fed, and customers etc.) in either direction.

Banks don’t really store money, you give them money and they give you bank credit, and then later on you exchange that credit back for money or transfer credit to someone at another bank and the two banks settle that with some asset (could be reserves etc.).

I think this is the main confusion, of course the physical cash the bank gives you for a withdraw might have been given to the bank during a deposit transaction, but in terms of the transactions on the balance sheet it did not come from there.

No, they say "I'm going to claim you can have $5000, because I can document that once I took in $1000 and I assert that I will take in $1000 again, perhaps many times".

Never mind $500, they don't even need to touch the $1000 you put in. They are just making it up as they go because they're a bank and can credibly assert that they will be handling capital in future.

They are using the FACT of having deposits to get the credibility to make up whatever loans they want. This is why leverage is like ten to one making the amount of capital in the world many times the assets of the world and everything in it.

When you withdraw $500 _they_just_make_it_up_.

It's a little more than a technicality. It's to do with the order of processes.

Essentially there is nowhere in a modern bank where they go "if deposits <= 0 deny loan"

Yet I see that time and again in economic models that supposedly have banks in them.

Bank lending and the treasury backfill process that manages the equity buffer are parallel processes that are linked by price not quantity.

That means that banks will and can lend to any and all creditworthy customers they get prepared to pay the current price of money.

It's a dynamic system that can expand and shrink based upon the demand for credit.

I'm not sure what you mean by "modern bank" - when is the cut off date for you?

I can assure you that banks can and do end up in a position where they are unable to provide loans. Admittedly the process isn't "if deposits <= 0" because banks don't just access funds from depositors. A retail bank will typically have a blend of funding sources where deposit accounts make up 30-40% of the liabilities on the balance sheets. The rest of the funds are raised on money markets and through inter-bank lending or bond issuance.

A bank's balance sheet looks like: - Liabilities: deposit accounts, issued bonds, borrowings - Assets: loans, reserves, equity.

I can assure you that "if reserves < X" will mean that a bank is unable to offer a loan to a customer - credit worthy or not.

I experienced a banking crisis first hand in the country of my birth and this is exactly what happened. This is why banking crisis are damaging to the economy as a whole - even creditworthy customers and businesses are starved of funds.

What do you buy capital Bonds/ equity with?

Bank deposits. The more deposits there are the more demand there is for bank capital in aggregate.

The system reaches an equilibrium at a price not a quantity. The quantity floats as required by the price of money.

Reserves are always available since the other banks lend them out and the central bank as lender of last resort. They have no control function in quantity. Only in their price.

You can buy bonds with credit, and that is limited by risk, as you say.

That's why banks like deposits, they are cheap, and thus they can lend out cheaper.

According to GP's linked article it's true when looking at the banking sector as a whole. For a single bank it seems like more of a technicality.

Also from the article:

> For example, if all of the deposits that a bank held were in the form of instant access accounts, such as current accounts, then the bank might run the risk of lots of these deposits being withdrawn in a short period of time. Because banks tend to lend for periods of many months or years, the bank may not be able to repay all of those deposits — it would face a great deal of liquidity risk.

Withdrawals go elsewhere and the central bank system then lends them back to the original source. In aggregate you just end up with a lot of label swapping. All that changes is the cost of the liabilities.

If I swap my labels wrong I can still go bankrupt though.

Deposits have two sides on the books: liability to the depositor, and cash as asset. They don't lend out the liability, they lend out the cash. That converts one asset, cash, into another asset, loans. I don't see this as misleading to say that they lend out deposits.

Yes, in a more realistic situation, they'll be using their assets as security for borrowing short term money that they lend out long term, so it's not exactly the same money, but it's still involved in the same risks.

Nope, that’s not quite it, because if somebody puts $10 in the bank, it creates $10 of assets (cash or reserves) and $10 of liabilities (deposits) so it doesn’t give the bank anything they can do anything with, apart from liquidity.

The only way it can work is if writing a mortgage creates an asset (the loan), and then a new liability is made to match it (a deposit of the same amount).

The security is 100% the bank’s capital, since any assets (cash/reserves) from deposits have matching liabilities.

The article I linked above from the UK’s central bank is confirming that all modem banks work the way I am saying.

It's been a long time since I did eco101, but I thought the multiplier was what made it possible for a bank to receive a $10 deposit and loan out $100.

The money multiplier has always been a simplified model that doesn’t actually explain how any modern real-world banks work.

When someone puts $10 in a bank, it most certainly gives the bank something they can do something with. In this case they can lend $9 out - assuming 10% reserve. On the balance sheet the bank then has $10 on the liability side - the deposit - and $9 loan and $1 reserve on the asset side. I think that you're confusing accounting with business.

That’s called the ‘money multiplier model’, it’s an incorrect model that doesn’t actually describe how any modern bank works. It’s basically a myth at this point.

I’m not confusing anything. Believe it or not, banks actually have to follow double entry accounting rules, and follow regulations about capital adequacy. Most countries don’t even have a reserve requirement, deposits generally don’t factor into lending at all.

Bizarrely, all the textbooks teach that that's how banking works, but it hasn't been true for decades.

It's characteristic of economics that it makes a song and dance about summing a geometric series while completely missing the point about how the system actually works.

What exactly isn't true in my statement?

- banks take in cash in from depositors (and other creditors)

- banks lend out cash to borrowers

- banks do not lend out all the cash that they've borrowed - they keep some in reserve

Do you have the name of any retail bank which does not operate according to this model? What does their balance sheet look like? The "textbooks" reference is not relevant here - my experience is from working in retail banking operations.

There is a academic paper here explaining how it works.

https://www.sciencedirect.com/science/article/pii/S105752191...

Essentially the bank can make the loan with a zero balance sheet.

If you work at the back end of a retail bank operation then you'll know that the loan department and the treasury department do not talk to each other about quantities. They talk to each other about price. Other than that the two operations operate in parallel. 'Funding' is just swapping the liabilities around in the market to try and reduce the amount the bank pays out in interest.

“I find it difficult to understand how this flow of funds analysis validates an assertion of general principle. What is described is a special case where a highly leveraged bank could find itself temporarily insolvent, with a very limited window of exposure before the bank would be found guilty of criminal fraud.”

https://www.cooperative-individualism.org/dodson-edward_revi...

I think the problem with what you're saying is the ambiguous use of the word 'cash'. Deposits, reserves, and physical currency are all 'cash' in some sense, but here it helps to be precise.

Banks don't 'lend out cash', they create deposits. Deposits are liabilities for banks. When a bank takes on a deposit, it also takes on a corresponding asset: either a loan in its books, or an increase in its reserve account.

In practice the reserve ratio is unimportant for a bank: in Canada and the UK the ratio is 0%. The real constraint on a banks ability to create money is it's leverage ratio, i.e. (assets - liabilities)/assets. In reality the formula is more complicated and uses risk weighting by asset class, but that's the crux of it.

For those wondering, here is an empirical paper on the subject demonstrating no connection between deposits and loans.

"This is the contribution of the present paper. An empirical test is conducted, whereby money is borrowed from a cooperating bank, while its internal records are being monitored, to establish whether in the process of making the loan available to the borrower, the bank transfers these funds from other accounts within or outside the bank, or whether they are newly created. This study establishes for the first time empirically that banks individually create money out of nothing. The money supply is created as ‘fairy dust’ produced by the banks individually, "out of thin air"."

https://www.sciencedirect.com/science/article/pii/S105752191...

Money is fungible so it doesn’t really matter where they’re getting the money from. They certainly aren’t keeping those deposits in a vault somewhere.

The accounting view is a useful frame, but money is fungible so there’s a real sense it which that view doesn’t represent what is “really” happening. See, for example, the MMT take on government financial flows.

Reply on news.ycombinator.com