A linguistic glitch tricks us into thinking bank deposits are deposited in banks | Hacker News Reader

A linguistic glitch tricks us into thinking bank deposits are deposited in banks

alteredstatesof.money

34 pointsjpkoning73 comments

This feels like a bizarre article that seems to be 'fixing' a confusion no one actually has. Also, if the premise of your article is that the 'dictionary is wrong because I feel otherwise and that's why everyone is confused', then you should consider the parsimonious explanation that no one is actually confused, that the dictionary is right, and that it is you has invented a confusion.

People do have this confusion all the time though: they think the bank lends out the money they put in, when really what’s happening is money creation as described in the article.

Interestingly, money creation is done by making loans. But nobody mentions the converse: money destruction is done by paying off the loans!

Fractional reserve banking is an amazing invention.

I like to use the analogy of malloc() and free() to communicate the concept to programmers. Commercial banks malloc() and free() deposits and the Treasury/Fed tag team malloc() and free() reserves.

To elaborate, it's interesting to observe that a non-zero interest rate effectively creates a resource leak. Since the deposits created are equal to the loan's face value and don't include interest payments, in aggregate there are not enough deposits available for every borrower to pay off their loans with interest. In other words, interest makes a non-zero bankruptcy rate a mathematical certainty.

What? Isn’t the money supply irreversibly expanded when a loan is created?

How does paying of the loan decrease the amount of money there exists?

When I loan 1000 the bank basically just writes a check for 1000 that is not attached to any money what so ever. When I’ve payed off my debt, the bank will have received 1100 and my understanding is not that they will simply delete the original 1000 and keep the change.

Paying off a loan contracts the bank’s balance sheet and the liability side thereof is what most private persons use as money.

They do just delete the original 1000. Otherwise they wouldn't be able to loan it out again, due to the reserve requirement.

But the two are independent ideas. You don't need lending or money creation for a bank to work, though even there the analogy is sound. The bank does lend money out and if a lot of people try to withdraw it at the same time, you get bank runs and collapses.

That being said, in principle, it is possible to have a bank that literally takes the cash you give it and puts it in a giant vault, and when you come to ask for withdraw it, retrieves it and hands it back over. It is literally the idea that the author seems to be railing against. You come with your money (water) and deposit (pour) it into the bank's vault (glass).

What exactly is the confusion?

The confusion is that this isn't how banks work. But most people believe that this is how banks work.

More than that, most people have no idea how lending really works. They believe the bank is literally lending their savings out. So if no one saves, there's no lending.

Even some economists believe a more complex version of this.

https://larspsyll.wordpress.com/2014/09/21/the-loanable-fund...

Well at a micro level, it is still what happens, the bank doesn't make money appear by magic.

The multiplier effect just happens because the original amount of cash has been deposited, then lent, then deposited again. So you have the same amount of cash in the system but two deposits, one backed by a loan, the other by cash, and the deposits are treated as "like cash" whereas they are mere IOUs backed by a financial asset.

A reserve (or government cash deposit) is not technically required for a bank to issue a loan.

Its _risky_, since it could lead to a situation where someone tries to cash out and you have nothing to give them, which would destroy your business and reputation as a banker.

But if you could make perfect loan decisions and no one comes to redeem their deposit, you wouldn’t need to have any reserves at all. Most regulators require a minimum amount of reserves to prevent instability or bank runs (and to control money supply)

That’s the difference, you’re not lending out the original amount of cash, you’re using it as a reserve in case someone asks for their money.

> in principle, it is possible to have a bank that literally takes the cash you give it and puts it in a giant vault, and when you come to ask for withdraw it, retrieves it and hands it back over.

That is indeed how banks originally worked (except gold, not cash). Then banks discovered they could hand out receipts instead of the gold, and people traded the receipts (how "banknotes" came about). Then banks realized they could issue more receipts than they had gold on deposit, and fractional reserve banking was invented.

The confusion is that modern commercial banks don’t work like a giant vault. So applying the analogy you described to, say, Wells Fargo, would mislead you about how Wells Fargo works.

You could have a bank that does what you say, but it wouldn’t be able to survive as a commercial institution.

Well said.

Most regular folks don't think like to think about banking, just like how they don't like to think about how their car engine works.

But if you ask them to describe how banking works, they'll inevitably resort to the some version of the "giant vault" analogy. Of course, that's the wrong way to think about banking. Which the article rightly attempts to dispel.

That's not a confusion but rather an incomplete (but, for most purposes, good enough) understanding of how banks work.

The money is lent out, though, in the specific sense that a bank run during a crash of the economy can† result in the bank not having the cash-on-hand to give back to its depositors.

† Or, at least, this was true in the US until 2007, after which savings banks were legally limited in their ability to leverage their deposits.

This is a problem of not having enough reserves, and not quite of lending deposits. Banks encourage people to put their money in the bank in exchange for deposits so that they have these reserves.

It’s like the article says: banks give out deposits in exchange for Government Cash, or for Valuable Long Term Promises. A run can kill a bank if they read the market wrong and can’t generate enough government cash to redeem the deposits.

Exactly, this is the perfect response to contrarian articles that insist on using vocabulary in strange ways.

People love feeling smart by reinterpreting normal questions like they are trick questions. There's a market for contrarianism that purports to overturn myths and conventional wisdom, and so people end up reinterpreting reasonable things as 'myths', and 'correcting' them with fun journeys into contrarianism.

I hope we can get to a point where we can recognize these articles as a familiar trope.

Indeed - it's mumbo jumbo.

A deposit in a bank is an asset for the holder of the deposit but a liability for the bank.

From this simple fact, the author ties themselves in linguistic knots of their own confusion.

It's not mumbo jumbo though, it's the minutiae of banking and it's important.

The hypothetical you have posited above doesn't actually reflect how it works. The actual money in the bank is an asset and is owned by the bank. The deposits (claims against that money by it's clients) are the liability. This is the difference, which is important to understanding how banks work, that is misses up thread.

You are not lending your money to the bank to hold. They are selling you a future claim on some amount of money from them, possibly (hopefully?) plus some interest against that claim. You will note that the FDIC, when talking about deposit insurance, uses the terms principal and interest. That's because the thing they are insuring is the loan contract between you and the bank.

You'll want the "How Bank Deposits Work" section of the below investopedia link.

https://www.investopedia.com/terms/b/bank-deposits.asp

https://www.fdic.gov/deposit/deposits/faq.html

It's similar to the confusion people create with the shorthand "I bought a song from iTunes". No, you bought a license to use that song in a specific set of ways.

Are you really claiming that a deposit account is NOT an asset for its owner or that a bank's deposit book is NOT considered a liability? That's what "doesn't actually reflect how it works" is suggesting to me?

Yet to contest these two simple facts makes no sense since they are trivial to verify?

Do you own any deposit accounts? Do you consider the money in them to be counted among your assets?

And if you look at Bank of America's balance sheet (https://finance.yahoo.com/quote/BAC/balance-sheet/), and drill into the liabilities list, you'll find the value of all their deposit accounts.

You should read what I wrote to try and understand it, as opposed to looking for what you expect to see. As I said, from the banks perspective they have an asset (the cash) and an offsetting liability (the deposit account). From the account holders perspective, there is an of course an asset (the title to the account).

It's slightly more complicated, as the bank doesn't actually hold your cash as cash. It deploys it in a variety of financial instruments to make money. Go back and read the assets section of that balance sheet more carefully, and you'll see this.

The point is that deposit is a technical term with a specific meaning. Namely, it is the agreement between the depositor and the bank that the bank now owes the depositor an amount of money which happens to be equal to the amount of principal deposited plus any accrued interest. If you had any deposit accounts and had thought about how they work, I'm sure you would have worked this out.

"The point is that deposit is a technical term with a specific meaning. Namely, it is the agreement between the depositor and the bank that the bank now owes the depositor an amount of money which happens to be equal to the amount of principal deposited plus any accrued interest."

I think my point and that of other commenters is that this is self-evident to almost everyone. It is not a confusion, because people's mental model of a bank (I give them money to hold on my behalf, and I can get that money back from them later, and depending on the situation, maybe a little more called interest) actually does represent this reality. Of course it doesn't represent it using the same terms because the terms are technical terms, but it's not clear how the lay person's mental model is wrong.

If I have understood the article correctly, I’m lead to disagree with what you’ve written. I think “most people” believe they get back the same kind of money they pay in. So, if they paid in state money, they’d get state money back. The article says they get “bank money” back.

If I go to the bank with a hundred dollar bill, 'deposit' it, whereby my account now shows $100 more, and then go to the bank after a year and withdraw $100 from my account, so that I now end up with a different $100 bill, it's not clear to me how the first $100 bill was state money but the second one was bank money.

A deposit is not an asset for the bank - it’s a liability - have you ever even looked at the balance sheet of a bank? I even provided a link which demonstrates that deposit accounts are listed as bank liabilities. is There is simply no question on this fact. So to put it simply you’re just wrong.

You’re confusing the concept of a deposit with the funds used to create the account. People who don’t understand basic double entry accounting or banking operations often make this mistake - like the author of the original piece.

> A deposit is not an asset for the bank - it’s a liability

You are repeating the words and ideas of the person you are responding to, yet claiming they have said the opposite and accusing them of not understanding what they wrote. I don’t understand your contribution.

I didn't find this article very enlightening.

It uses too many words to explain a straightforward concept: You hand your cash over to a bank, in exchange they give you an IOU. They give out IOU's far in excess of the amount of cash assets they actually hold.

The concept of "short-term promises in exchange for long-term promises" was a bit more illuminating.

It might also be misleading to say the bank takes ownership of your deposits. More accurate to say they take custody, given the strong fiduciary obligations the transaction imposes on them (at least in developed economies).

The intrusive sign-up-to-subscribe form a mere few paragraphs in doesn't win the author any love from me, and while I don't generally mind mspaint-flavored art I found the annotated illustrations somewhat amateur. The whole article feels like it was written by a youth who just discovered the concept of fractional reserve banking and wants to educate the world.

Just to be clear, the bank holds more assets than its liabilities, otherwise it would have gone bankrupt. It just happens that a lot of these assets are themselves IOU from private customers, usually backed by a property asset.

Thanks, great point. I updated my remark to say "cash assets" and I'm open to better phrasing.

typo:

s/bankrupt/insolvent/

As you're probably aware, bankrupt is not a relevant concept for banks. Creditors- depositors- do not get a haircut, their deposits are insured.

No, banks can go bankrupt like any other company. I think you are refering to Chapter 11 (in the US), which is true, isn't practical for a bank, because the bank has no other business than being trustworthy to its creditors, so it couldn't really operate under Chapter 11.

The deposit insurance only applies to deposits below a certain threshold and doesn't concern the bank, it just means the state will make the customer good for its loss in a bankruptcy.

Now in practice central banks will do everything they can to avoid getting there, first by forcing banks to hold a lot of capital and liquidity. But it can still happen, and one of the tools available is bail-in, which effectively replicates the effect of a chapter 11 over a week end, outside of courts.

I think in most jurisdictions, wholesale creditors are the most likely to get a haircut in a bail-in, and the insured part of the deposits is explicitely non-bailinable. But any deposit above the insured threshold can in theory be affected by a bail-in and have a haircut applied.

But a lot of fail safe mechanisms will have failed before that happens.

Thanks, yup- to be more precise, a state of insolvency- liabilities > assets at close of business- may result in one or more outcomes, one of which is essentially bankruptcy + bail in, but another might be, eg being absorbed/acquired?

Correct. Though just to be precise, a bailin would happen well ahead of a bankruptcy, the regulator wouldn’t wait for the situation to be that bad to call the bank non viable (except in a jump to default scenario).

Absorbed/acquired: that’s a bailout. Always an option but I think everyone agrees no one wants to see that happen again unless it is a willing private buyer.

It is a linguistic trick. Same as "credit card" (a loan card), credit implies the opposite.

No. Credit, as in implied future payment, literally does mean loan, and another meaning of the word (opposite of debit) is also what happens when you use a credit card - an intermediate credits the merchant account so you don't need a separate line of credit for each one.

You're making the same not-even-faulty assumption as the article author, that a word with several related meanings is somehow wrong or a trick or at all tough for the average person to comprehend.

Sure it does, like many English words that have different meanings. And a marketer will choose the word with the fewest negative connotations.

You can use a credit card as a loan card, sure, by taking cash out an ATM or not paying your entire balance at the end of the month. And yes, loads of people obviously do that, and that's a trick.

But can you really say it's a sleight of hand that soda isn't called 'liquid obesity, period.' by its manufacturers even though that's the result if used to the extent they'd want you to?

No, because soda doesn't have another meaning.

This also means you've failed to make a distinction between the two forms of money in our society: state money ('water'), and bank money ('promise for water').

Technically the "state money" is also just a promise. In years long past, it was a promise of a particular quantity of precious metal. More recently, it's not a promise of that, but they strongly imply that one who possesses state money can use it to pay taxes to the state.

You can also use state money to settle debts established by the courts. Money is not an illusion.

FRNs are not an illusion, just their intrinsic value. Originally, dollars were exchangeable for gold or silver. Now, we get the privilege of paying taxes and court fees with FRNs. Outside of those arbitrary fees, they have no value. Its current currency value is that everybody (in the US) takes them and most international markets are based on them. once the international markets switch to some other currency, the FRN will lose a rapid part of its value, and the real brunt of the inflationary principles the Fed has adopted will be felt. It will be quite painful.

There seems to be this meme that hyperinflation is always painful. Long-term hyperinflation is painful because nobody wants your currency and that causes all kinds of problems.

Short-term hyperinflation is short-term disruptive, but it also has the result of effectively wiping out your debts (which would be helpful on net to most in the US, including the government), and then people just adjust to the new prices, which are higher but become stable, and wages rise to compensate.

The main loss to the US from not being the reserve currency would be that they couldn't keep printing even more money without incurring the normal amount of inflation that usually implies for everybody else. But that's assuming the rest of the world is even interested in handing that power to somebody else. And who would that be? Everybody wants it to be themselves, which is what nobody else wants. Meanwhile all the powerful international holders of US debt have a huge interest in it continuing to be the US, since they're the ones the wiping out of dollar-denominated debts would hurt the most.

And none of that seems especially likely in the immediate future, because the Fed is doing all it can right now to prevent deflation, by keeping interest rates on the floor and printing a ton of money. It would be so easy for them to prevent inflation right now that all they would have to do is stop actively doing half the things they're doing to prevent its opposite. So it only happens if they want it to.

> once the international markets switch to some other currency, the FRN will lose a rapid part of its value

I mean, other countries manage to have inflationary monetary policies, without their currencies being the global reserve currency.

Lots of valid critique for this piece. Will add that "state money" and "bank money" is a poor way to talk about credit.

In terms of references, tons and tons and tons of prior art defining money and so forth. A book I recently enjoyed a great deal is:

The Nature Of Money, Geoffrey Ingham

https://www.amazon.com/Nature-Money-Geoffrey-Ingham/dp/07456...

And this more recent Bank of England paper is exceptional:

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

The point of this article is: a bank deposit is an asset for you, the depositor, and therefore a liability for the bank. A mortgage is a liability for you, and therefore an asset to the bank.

Sometimes we get confused and we think that the bank deposits must be assets for the bank, and mortgages must be liabilities. This isn't because of some linguistic quirk. It's because its good for the bank to have a lot of deposits, because it means it's doing a lot of business. In the same way a company with a billion-dollar line of credit is likely doing better than a firm who can't borrow a cent.

Also, when the bank originates a loan it also creates a corresponding deposit. Therefore a bank having more deposits indirectly indicates it’s originating more loans.

It's the other way round. The first part isn't true - the bank doesn't buttonhole some saver into giving them a few hundred grand when you take out a mortgage. (It might get the deed to your house, physically or metaphorically - but there are also unsecured loans).

The second part is true: when you make a deposit the bank has some more cash and will surely invest it/loan it out/deposit it at another institution. Those three phrases all mean roughly the same thing: the bank will take the pile of copper and linen you gave it and use it to purchase another asset.

I wrote nothing about savers. Both the loan (bank asset/customer liability) and the deposit (bank liability/customer asset) are created from nothing at the same time when the bank originates a loan.

The loanable funds model, which it appears you are alluding to, is observably an operationally incorrect fantasy.

> "If this is you, you've made an error, and have fallen into the trap of thinking that banks 'lend out the water that I've put inside them'. This also means you've failed to make a distinction between the two forms of money in our society: state money ('water'), and bank money ('promise for water'). And both of these misunderstandings are facilitated by the linguistic glitch we are trying to deprogram."

I don't think this clarifies much. Talking about categories like "state money" and "bank money" doesn't make the problem clear.

I have come to the conclusion that the problem with fractionally reserved money isn't the fractional reservation mechanism at all. Rather, the problem is that banks effectively promise the same money to multiple people at the same time. There is one "piece" of money, but multiple people have claims on it.

If deposits were timed (like CDs) and the banks could only loan out the money for the time it was tied up, there would be no problem. In fact, there would be no need for a reserve ratio, just loss provisions. A "piece" of money could be loaned out an infinite number of times, in fact. Every bank would just need to show a loanable amount curve over time as well as a loaned amount over time, and you could easily see if a bank was in trouble or not well in advance.

Demand deposits could not be loaned, of course. Basically, force banks to borrow long and lend short.

This is what I'd like to see tried.

> I have come to the conclusion that the problem with fractionally reserved money isn't the fractional reservation mechanism at all. Rather, the problem is that banks effectively promise the same money to multiple people at the same time. There is one "piece" of money, but multiple people have claims on it.

Banks get overnight loans to cover the case where creditors want more money than they currently have access to.

There is no single piece of money anywhere anymore.

When you deposit a $100 federal reserve note in the bank you are only giving that bank a means to pay taxes or pay back loans from Federal Reserve Banks. If you take a $100 note to a federal reserve bank they will look at you funny.

The reason this works is that U.S. currency is so stable that any potential commodity you might desire as currency can be paid for with USD money from virtually any bank in the world. Banks are just an accounting system at this point who barely hold any physical thing of value, and only on a just-in-time delivery model. The rest of the economy satisfies the market for transactions between money and goods.

Why doesn't it all fall apart? The federal reserve raises interest rates when there is too much money in circulation, urging banks to pay off their liabilities, thus destroying money.

Central banks already ensure that if multiple people the money is promised to want to make withdrawals at the same time, they can all have the cash.

But if a mortgage requires people to lock away funds for quarter of a century (because the bank can only loan out the funds if someone has the money tied up for that long) then whoever's lending is [i] diverting the funds away from more productive activity [ii] going to expect a very high interest rate because they don't see their money for such a long time and [iii] likely to be richer on average than the current beneficiaries of banking activity. Reducing transfer from productive to unproductive activity and the scale of transfers from poor to rich are desirable: maturity transformation is therefore a feature of the system, not a bug

[n.b. if you force banks to borrow long and lend short you've eliminated the fractional reserve mechanism anyway]

The "bank" in "bank deposits" is an attributive noun (a.k.a. a "noun adjunct"). In "oil deposit," since oil is a mass noun, we interpret the phrase as meaning "a deposit made of oil." But "bank" is a count noun--the phrase "a deposit made of bank" makes no sense--so the natural interpretation is "a deposit related to a bank."

Essentially, "bank" is just a noun being used as an adjective (as in phrases like "horse race," "corn maze," "motor vehicle," "chicken noodle soup bowl," etc.)

This speech by the SNB chairman describes fractional reserve banking well in my opinion: https://www.snb.ch/en/mmr/speeches/id/ref_20180116_tjn

It sounds like he's saying that the banks should have a type system for money, maybe something like this:

  datatype Money 
  = Deposit of int
  | Reserves of int

(Or maybe a unit system) and we shouldn't think of these as comparable..

You just invented T-accounting!

Lol - I wish, but the datastructure I posited is a discriminated union, not a tuple.

What is T-accounting?

The basic concept behind accounting, the T represents the balance sheet which must be balanced at all time, and every change in the balance sheet is always a tuple of two operations, either on both sides of the balance sheet, or on two lines of the same side.

So if you start with an empty balance sheet and a customer comes with cash to deposit, your first operation is: increase deposits | increase cash.

Then the bank lends money to a borrower: decrease cash | increase loans to customers.

Then the borrower pays an interest: increase cash | increase equity

Then you pay some interest to the depositor: decrease cash | decrease equity

etc.

Also called double-entry bookkeeping: https://en.wikipedia.org/wiki/Double-entry_bookkeeping

It's not a linguistic trick. It used to be that banks had to keep a certain percentage of they money they had on deposit on the premises. Slowly, this amount (the "reserve rate) was lowered to 0% under Obama.

>Slowly, this amount (the "reserve rate) was lowered to 0% under Obama.

Source for this? AFAIK it was done this year, under trump not obama.

>As of March 2020, the minimum reserve requirement for all deposit institutions was abolished, or more technically, fixed to zero percent of eligible deposits. The Board previously mandated a zero reserve requirement for banks with eligible deposits up to $16 million, 3% for banks up to $122.3 million, and 10% thereafter. The removal of reserve requirements followed the Federal Reserve's shift to an "ample-reserves" system, in which the Federal Reserve Banks pay member banks interest on reserves that they keep in excess of the required amount

https://en.wikipedia.org/wiki/Reserve_requirement#United_Sta...

You two are mixing up required reserves ratio:

https://tradingeconomics.com/china/cash-reserve-ratio

and a bunch of other yields, which connect to the the rate paid on reserves, the rate on excess deposits, fed funds, libor etc etc.

https://tradingeconomics.com/united-states/interest-rate

Wait, I thought the reserve ratio was set to zero for the first time this year, because of the pandemic?

It used to be that banks had a regulatory requirement to do it. That's very different from the mechanistic requirement most people envision, where it's just not possible for a bank to make a loan if they don't have enough money stashed in their vault.

“Reserves” are a completely different thing from “deposits” though. I don’t think the bank is deliberately being tricky, but I think the idea of banks “lending out people’s deposits” is not quite true.

The main mistake is thinking deposits are taken out of banks.

Once you realise that a bank note is a receipt for a deposit at the central bank it all becomes clear.

We just swap deposits between ourselves

In Ireland, you make a lodgement when you are paying into a bank.

I was in disbelief when I first realised that no-one in the UK knows what a lodgement is.

Eh, this sounds more like linguistic overengineering really. And that accounting example is wrong.

I can’t tell whether this is debunking a myth I didn’t believe to begin with, or if I do believe the myth but am too dumb to recognize that it’s what this article is about. This is bizarrely opaque for so many words about a fairly small (I think?) scope of knowledge.

I’m also not quite sure what the point of this was. It’s seems a long winded way to say that the money you deposit is not physically there regarding absolute ownership.

That or I too am completely missing the point.

Why even use the word "glitch"?

Reply on news.ycombinator.com