Hacker News new | past | comments | ask | show | jobs | submit login
FDIC Takes over Silicon Valley Bank (fdic.gov)
3172 points by khuey on March 10, 2023 | hide | past | favorite | 1576 comments



An explainer post [1] connected to that Tweet is something I found extremely informative (assuming it's accurate):

"- In 2021 SVB saw a mass influx in deposits, which jumped from $61.76bn at the end of 2019 to $189.20bn at the end of 2021.

- As deposits grew, SVB could not grow their loan book fast enough to generate the yield they wanted to see on this capital. As a result, they purchased a large amount (over $80bn!) in mortgage backed securities (MBS) with these deposits for their hold-to-maturity (HTM) portfolio.

- 97% of these MBS were 10+ year duration, with a weighted average yield of 1.56%.

- The issue is that as the Fed raised interest rates in 2022 and continued to do so through 2023, the value of SVB’s MBS plummeted. This is because investors can now purchase long-duration "risk-free" bonds from the Fed at a 2.5x higher yield.

- This is not a liquidity issue as long as SVB maintains their deposits, since these securities will pay out more than they cost eventually.

- However, yesterday afternoon, SVB announced that they had sold $21bn of their Available For Sale (AFS) securities at a $1.8bn loss, and were raising another $2.25bn in equity and debt. This came as a surprise to investors, who were under the impression that SVB had enough liquidity to avoid selling their AFS portfolio."

[1] - https://twitter.com/jamiequint/status/1633956163565002752


I have some family who (with some other partners) founded a small community bank that has grown over the years.

They expanded in some areas by buying other small community banks, specifically in areas where there was a big increase in income in the local area (from mineral rights, etc).

The smaller banks that they bought were in a situation where suddenly they had large amounts of cash incoming, and customers who were paying off / not taking out loans like they used to.

They didn’t have the reach (mostly confined to a small rural region) to use that cash to give out loans elsewhere so they looked to merge or be bought by someone who did.

Until I heard about those banks I hadn’t considered “too much money” was a problem.


At first glance, bank balance sheets are unintuitive and feel 'the wrong way round'. When someone deposits $1m at a bank, the bank doesn't have $1m more assets, it has $1m more liabilities.

(Yes, this is a gross over-simplification)


To my mind, although it's in-principle equivalent, the clearer way to think about this is that banks borrow money from depositors and lend that money via loans or investments.

The primary business of a bank is borrowing short and lending long - where short and long refer to the holding time: i.e. taking demand or short-duration term deposits and making mortgage, car and other types of loans.

If you do this badly, you can lose money due to duration risk (you might end up paying more on your demand deposits than your mortgage book is bringing in), but you also have liquidity risks because your depositors can ask for those deposits back faster than you unwind your lending.

If you have both of these occurring at the same time, you're then in severe difficulty, because the only way to repay your depositors is by borrowing money ... which is going to be harder if you look unprofitable ... and that very borrowing can exacerbate the perception of a bank in trouble.

That's basically what happened here.


That’s not how it works. This is a good primer by the Bank of England: https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...

Bank lending always creates new money. They don’t (and can’t) “lend deposits”.


OK, sure. It’s a distinction without a difference though. If you read pg. 5 of that document, it explains how the effect of market forces results in essentially the same net effect for an individual bank as if they did “lend deposits”.

This, plus the other effects (like what the BoE calls “prudential regulation”) are the reason why you don’t see banks with zero deposits and a trillion pounds of loans.

Credit vs quantity theory of money is basically a concern of monetary policy (like: what is the expected effect of quantitative easing?) and doesn’t really bear on decisions at the level of individual banks or borrowers: things look consistent with both theories.

Like: it doesn’t matter whether classical or relativistic mechanics are “true” if you’re only following the flight of a baseball.


Maybe, but in the case of SVB, deposit amounts were suspiciously similar to assets....

The BOE is simply pointing out in the link that they are the only proper bank in the UK. All the others borrow from the BOE to make loans. Commercial banks don't "create" money any more than the BOE "borrows" it from elsewhere.... The BOE creates money and lends it to commercial banks who put up a deposit and make a reasonable case for credit.


In a fractional reserve system the amount they can loan out is a multiple of their deposits, with the excess money coming from the central bank.

Also why link an article from the UK when the subject is a US bank? There are differences between the two countries banking systems and monetary policies.


Cash is fungible so the difference is not important. A bank has to balance its books eventually.


Of course banks are lending deposits. They're throwing in some investor money too but vast majority of loans they write is using deposits. You can try to make some sort of "number on screen go up" argument, btu the underlying value comes from somewhere. When someone takes a loan and withdraws dollars they're usually getting money that was deposited by a customer.


That's a misunderstanding - when a bank originates a loan:

- the bank records an asset (the loan) and a liability (the deposit in the borrower's account) - the borrower records an asset (the loan money now deposited at the bank) and a liability (the loan)

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

At this point the money literally came out of thin air, there's more total currency in the system then there was before.

It might seem like an irrelevant nuance but it's pretty important to understanding how the money supply works


But is it real? It’s more like an IOU, right? They don’t reduce the amount of money in the depositor’s account, but if the loan defaults would they be able to return the money to the depositor? My guess is no, not really (assuming no other outside cash sources).

So even though the amount of currency looks like it has increased, it hasn’t really… unless I’m misunderstanding still. But my impression is this is pretty much what happened yesterday with SVB.


Your misunderstanding is around what money really is: it's just an accounting unit. It's literally a number in a database somewhere.

It's very hard for people to accept this because of how most people basically trade 1/3 of their lives to get some. It just doesn't "feel right"

Currency is just a physical representation of that unit.


Am I wrong though that if the loan were to default and the depositor tried to withdraw all 1000, the bank wouldn’t be able to return it?

Perhaps because of the insurance from central banks this problem is avoided.


That's wrong. The loan amount was never anything but a number that someone at the bank punched into a database. It doesn't have anything directly to do with someone else withdrawing their own money.

If the bank is giving cash to the withdrawer, that would come out of their capital reserve I guess. If they didn't have that, they FDIC insurance would kick in.


You’re wrong. If someone took out a loan and withdrew all the money before defaulting the bank would not be able to pay deposits. Insured deposits would be paid by the government and uninsured deposits would go poof.

You can say the loan is a number in a database, but once it’s withdrawn from the bank the bank can’t just go into the database and undo the loan.


Your understanding of banks taking in deposits then lending them out hasn't been true for decades. A bank with no deposits can make loans just fine. They can just borrow at a lower rate than they loan if they actually need to hand out cash.

Your example doesn't make much since because any bank has many good loans and a capital reserve to deal with the few bad loans.


This was basically what I was asking. Okay so, because the bank has access to an external supply of “money”, they can loan out more than they have in reserve. Thanks!


Yes, BUT they do not need to get 1000 USD from anywhere to put 1000 USD in your account. They literally just say you have 1000 USD. You can then electronically transfer that to anyone else's bank. The banks then borrow government money overnight to settle accounts. That's the only place government money comes into it.

Economists actually refer to the USD in your account as bank money because it did not come from the government. You could just as easily think of that money in your account as a bank issued stablecoin pegged to the dollar that's convertible to dollars.


> The banks then borrow government money overnight to settle accounts.

Maybe that’s already implied in your description but they don’t need to borrow from anyone if they had enough reserves to start with.


This is expensive. Banks can do this, but it's more of a last resort than the main plan.


Money is not "real" so you can consider this IOU is actually money. And if everyone agrees then it is.


> But is it real?

Yes. Money isn't 'real' in the sense people typically mean. It's an accounting unit, so it's exactly as real as a liter or a kilometer.


except people withdraw money and the bank needs to give those people actual cash. That's deposits. Of course they could take overnight loans or whatnot but deposits are cheaper.


The vast majority of money is not withdrawn. It's all transfers. And that is a major point and difference.

Also, if it was all withdrawn it's a bank run, which has not been a problem in the US for quite a while.

Overall it cancels out for the large banks, of course people transfer money they just borrowed from the bank away, but that bank also has customers that are recipients of money. That, and other mechanisms between banks and banks and central bank(s) to balance such things within all the banks in the overall economy. Money withdrawn from one bank goes to another one, and everybody is not just sender but also recipient from someone else.


I am responding to a post that said

> Bank lending always creates new money. They don’t (and can’t) “lend deposits”.

It's just not true. The banks do transfer literal cash from deposits out of the system through loans. The cash in banks didn't come from thin air.


Well, even reserve balances are created out of thin air… it’s all just accounting


> The cash in banks

Why do you keep insisting on the "cash"? The by far least important kind of money? First of all everything is just virtual, electronic. Only a tiny fraction - and only temporary! It's not stocked at home, but used to buy something and that means deposited again in another bank - is "cash".

And yes, the bank deposits are "out of thin air" -- https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m...

Yes banks need some reserves, but this is waayyyy more complicated than them using the deposits. Which, for the most part, are not cash. That's only that tiny fraction of paper money use din circulation, which also is a lot less than it used to be with all the card-money.

https://www.bankofengland.co.uk/explainers/what-is-money

In the UK, as an example:

> 96% is electronic

> 4% is cash

USD is different because that currency is used for far more than just regular circulation, and world-wide. But even then estimates are ca. 8% cash maximum, for the world, which includes a lot less sophisticated economies. A bit more for the USD for obvious reasons, but most of that is not circulating in the US economy.

In any case, even in the US, with all its cards and card payments, money is mostly electronic and not cash and is transferred from bank to bank and hardly leaves the banking system.

ABOUT THAT TERM "OUT OF THIN AIR"

I've seen articles arguing money is not create out of thin air, e.g. https://cepr.org/voxeu/columns/banks-do-not-create-money-out...

> commercial banks create private money by transforming an illiquid asset (the borrower’s future ability to repay) into a liquid one (bank deposits); they would quickly be insolvent otherwise

Uhm.. yes, exactly. They can't do it at will, they need a borrower to sign. I mean, now we start arguing definitions, always a bad sign, especially when both sides actually are in perfect agreement and it's about the words used. This article to me is just weird, arguing it's not out of thin air, while everything being described says just that, only that their understanding of "out of thin air" is subjectively different than that of many other people. But there is no difference in knowledge and opinion about the underlying mechanism, only disagreement about some fuzzy term that some like to use and some apparently hate for whatever reason.

For arguments about definitions I would like to point to "Disputing Definitions" (6 min read) https://www.lesswrong.com/posts/7X2j8HAkWdmMoS8PE/disputing-...


I’m not reading that, but I’m using “cash” because that’s what I’m talking about. Yes most of the loan money isn’t cash, but the cash has to be available for withdrawal so “it’s just numbers on a screen” doesn’t actually work all the way down.


Refusing to read about the thing you don't understand is pretty silly


Banks don't hold substantial cash. They send whatever cash they collect to the local federal reserve to be recycled and get fresh cash to hand out to people every morning from the same federal reserve. I don't think I've gotten a "used" bill from a bank in over a decade.


So let's assume a bank has collected deposits worth £1000 from all of its customers. Now the bank makes a loan of £100 to one of its customers and puts the money into the customer's current account at the bank.

According to your theory, what is the total amount of deposits at this bank after making the loan?

If, as you say, banks were lending deposits, then the total amount of deposits cannot possibly have changed and that new loan of £100 would have to come out of some other customer's bank account.

I would be pissed if my bank account balance suddenly dropped and the bank told me, sorry we have lent your money to someone else.

So this is clearly not what's happening.


They give your money to that guy. And he puts it in his account at the bank and they give it back to you. The money is still real though. Without the deposits they could not make any loans to begin with.

If someone comes to take out a loan for $1000 and withdraw the money what would happen? They'd go get the money everyone left there and give it to the new customer. Now they bank has no money because they just lent all the deposits. $1100 in deposits and $1100 in loans


“ Without the deposits they could not make any loans to begin with.”

Incorrect, banks create deposits when they issue loans and only require reserve balances sufficient to satisfy net flows of funds between institutions.

They can also borrow those reserves.

So deposits are required to increase profitability, and of course a minimum level of profitability is required to be solvent, but banks in now way “lend out deposits”.


If you actually believe this describes the process then follow me on this thought experiment.

TWB (unrelated to SVB) realizes that it’s in trouble - customers are taking a lot of their deposits out, and confidence in the institution is low.

So, instead of trying to liquidate some of their assets or raise capital (which is what would be conventional) they decide to lend their bank president a trillion dollars of interest free loan with a recall feature, on the contingency that he deposit it in a non-interest bearing account with a 300 year notice requirement.

In your understanding of the mechanics of banking; this is fine. The bank magics $1tn of deposits out of thin air, records an equivalent asset, and can then just pay all of their customers from these new deposits they have.

Or not? And if not, why not?


By issuing deposits, each bank creates essentially its own money which is valid as long as it is inside the bank (used in transactions with other clients that use the same bank), but when clients send money to other banks, the bank has to use 'central bank money'.

Which means banking system as a whole creates 'private money' when flows of 'central bank money' between banks are balanced, but each individual bank cannot do it faster than others, otherwise these flows will be negative and it will be losing 'central bank money' and/or hard assets (which would be sold to acquire 'central bank money').


In your worldview, where does the money a bank needs to operate from?

In your worldview, can you explain how it is possible for a bank to have far more loans on their books than they have deposits (see: fractional reserve banking)?


> In your worldview, where does the money a bank needs to operate from?

The answer in the US is simple: the Fed.

A bank can make an infinite amount of loans. They are not constrained by anything.

The only "limitation" they have is the future default rate of the people who accept loans from them.

The way a loan works is: they make a loan to someone and then "just in time", they borrow the money from the Fed to cover that loan.

Now, many banks take customer deposits. I think there are 2 reasons they do that:

1. Deposits may be less expensive than borrowing from the Fed.

2. Account holders are warm leads for future loans.

But in principle, a bank doesn't have to take any deposits.


> A bank can make an infinite amount of loans. They are not constrained by anything.

They are constrained by capital and/or reserve requirements

> The way a loan works is: they make a loan to someone and then "just in time", they borrow the money from the Fed to cover that loan.

Nope. Banks can borrow against their government securities from Central Banks by "rediscounting" them during discount windows, should they need extra liquidity. Note that this is not the preferred method since Central Banks usually charges a premium. These are secured loans.

See https://www.federalreserve.gov/monetarypolicy/discountrate.h...


>The way a loan works is: they make a loan to someone and then "just in time", they borrow the money from the Fed to cover that loan.

Ehhh, how does that work exactly? They go up to the discount window and borrow from the Fed every time they issue a loan?


> how does that work exactly?

I don't know. I don't work at a bank.

But loans aren't typically available immediately - they take time to clear. It doesn't seem unreasonable to me that they'd bundle the day's loans (or maybe a few hours at a very large bank) to limit the number of transactions they'd have to make.


The money that gets borrowed from the Fed is used to settle balances between banks. When a bank originates a loan, they just add a number to an account. The money doesn't come from anywhere. This works mainly because most money transfers are electronic.

tl;dr yes, banks create money out of thin air. It's been this way for decades.


> A bank can make an infinite amount of loans. They are not constrained by anything.

That's incorrect, there are very strict limits on the solvency ratio placed on the banks that you usually deal with.

https://www.investopedia.com/ask/answers/052515/how-are-risk...


> can you explain how it is possible for a bank to have far more loans on their books than they have deposits (see: fractional reserve banking)?

Fractional reserve banking is about having more deposits (on the liability side) than reserves (on the assets side).

I won't say that it's not possible for a bank to have far more loans on their books than they have deposits (they could finance themselves differently) but I'm not sure if that actually happens. Can you give an example of a bank having far more loans on their books than they have deposits?


Banks don’t have far more loans than deposits. When they make a loan most of the money is just redeposited, so effectively each oringal deposit gets 10x’d by people just leaving the money in the bank. That’s how fractional reserve banking and “creating money out of thin air” works.

If none of the loan money was redeposited than the bank couldn’t create new money.


You are thinking of a nation - and even a single bank - as a closed system. But this is an oversimplification that makes it impossible for you to make steps in your understanding of how this all works.

I've worked for a bank, both on the 'banking' side and on the IT side. One of the first things that gets drilled into your head is that banks create money. With every loan on the books more money gets put into circulation. There are some restrictions on how much you can put into circulation and there are some restrictions on how much cash you have to have on hand compared to the number of deposits that you have lying around.

But a bank could easily (as long as the bank is 'solvent' according to the rules set by the local central bank) write loans well in excess of it's deposits, technically it need not have any deposits at all.


I wonder if it’s correct to also think about it in this way: idle cash devalues over time. Coupled with interest (no matter how small) that they pay out to their depositors, this exposes banks to future liability. To counter that, they have to give “jobs” to as much of this cash as possible so that they can make those payments while also pocketing a profit for themselves.


This is indeed a reasonable way to think about it. A dollar tomorrow is worth less than a dollar today.

You might enjoy this light reading: https://www.investopedia.com/terms/n/npv.asp


There are plenty of instruments though with different maturity dates, from a single day to decades.

There is no reason to lock money in 10 year notes, it can stay as cash or go to money market or short term obligations.


Too much idle cash is a wasted asset.

As for money markets, don’t those get invested in treasury bills/notes anyway? Why go through a “middleman” when one has enough volume to invest directly?

Short-term obligations may not provide the yield that their financial structure requires.

10-year notes, in certain situations, provide an optimal combo of yield and risk. Provided, of course, that nothing major happens to the economy which wasn’t the case here. Then again, who’s good at predicting that?

In the end, it seemed like, given what was true at the time the decision was made, SVB made a rational choice.


> As for money markets, don’t those get invested in treasury bills/notes anyway

Yeah, but the market price basically stays at $1 and whenever they "break the buck" it's a huge deal.

10 year loans "break the buck" so often that it's extremely strange SVB didn't do anything about it.

Basically they should have invested in securities that were safer and fluctuated less, but they got greedy chasing yield and got found out.


> banks borrow money from depositors and lend that money via loans or investments.

Do they? Reserve requirements are almost non-existent. A better way to put it is that banks pretty much have a state-granted privilege for creating money and in the form of loans.


People read “reserve requirements have gone to zero” (which is true) and assume it is so that banks can keep lower reserves.

In fact, reserve requirements became meaningless because banks had (and continue to have) so much more in reserves than they were ever required to have under the prior system.


I don't understand this, could you clarify? If they have so much more money than the requirements demand, what is the point of lowering the requirements? Who does it help?


The point wasn’t that reserve requirements were “lowered”; they were removed - because the Fed moved to a different mechanism to set short-term interest rates.

If banks choose to keep lots of reserves (because you’re paying interest on them), it’s hard to change the willingness of banks to lend by changing the reserve requirement, right?

In economic terms, the supply of reserves is meeting the demand curve for reserves in an inelastic region.

So the Fed decided to announce a new approach; the so-called “ample reserves regime” where short term interest rate control would be achieved by administered rates rather than by the reserve requirements of the “limited reserves regime” that went before.

If this sounds very far from the quotidian business of deposit taking and loan making: that’s because it is.


And they can get reserves whenever they need to, because that's how the interest rate policy is effected.


I hear this a lot and it's in a lot of "explainer" videos on youtube and so on. The idea that banks have a special state-sponsored privilege to create credit is completely and very obviously untrue.

Try this thought experiment. You and I are in a bar. We order drinks but oh no the bar's card machine is down and you don't have cash. No problem I'll lend you a tenner.

1) Am I a bank?

2) Did I first contact the state to check it was ok to create money in the form of a loan to you?

I go back to the bar on another day on my own. Oh no! Their card machine is down again! This time I don't have cash on me. No problem the barman sees me all the time I can have the drink this time and pay up when I next get there and either have cash or their card machine is working.

3) Are they a bank?

4) Did they first contact the state to check it was ok to create money in the form of a loan to me?

The answer to all these questions is obviously no. Credit is created throughout the economy at all levels and it is not a special function perculiar to banks.


You are talking about credit, not money. Banks create actual state-backed money, which can be physically (or more likely electronically) paid to a third party after they've loaned it to you.

A better example would be if you were in the bar alone, and had no money, and instead just wrote an IOU on a bit of paper with your contact details. If we lived in a mythical 100% trustful utopia, then the barman would just accept this as money and so would anyone else in future who the barman needed to pay for anything. But we obviously don't live in such a society so unfortunately no, not anyone can just create money!


Banks don't create state-backed money either. The central bank creates money and gets it into the financial system by performing open market activities (eg buying treasuries with it) or depositing it into the accounts it holds with various banks.


No, the banks create money. If you've watched all these explainer videos surely you must understand basic FRB.


Fractional reserve banking absolutely does not create money[1]. It creates credit exactly the same as if you lend a friend of yours some of your money. You now have an asset (the loan) and your friend has a liability (the debt) and the amount of M1 or M2 money in supply has not changed.

The explainer videos I have seen are wildly wrong. I learned this stuff by reading the Basel accords [2], working with banking regulators and central bankers, working on banks' capital reserve models etc. That is to say I know how this actually works because I have been inside the sausage-making process for good or ill - I didn't learn it second-hand from someone who probably also learned it second-hand which is the feel I get from these videos.

Almost any time you see someone "explain" fractional reserve banking it is about 99% probably total bullshit. It's got to the point where "fractional reserve" is almost a trigger phrase for me - I know when I hear it that it is highly likely the speaker doesn't know what they are talking about. Almost like when you hear the word "fiat currency" you know it's very likely someone is going to try to shill you some crypto. An explanation which is not nonsense is here. [3] Fractional reserve banking means the bank doesn't need to keep the full amount of deposits in reserve, but can use some percentage to make loans. These loans are assets the bank has, but as with the example I gave above where you lend to a friend, no additional money is created in this process and when a bank does it, it's not fundamentally any different.

[1] You can find the definitions of M1 and M2 money and a good explanation of what's included in the definition of "money" here https://www.investopedia.com/terms/m/moneysupply.asp

[2] Which are not secret by the way - you don't have to be initiated into the templars or something to understand them. They are boring as all hell to read but otherwise reasonably understandable with a little bit of background https://www.bis.org/basel_framework/

[3] https://www.investopedia.com/terms/f/fractionalreservebankin...


I feel like there are quite a lot of people here that do not understand that the relationship between theories of money and the actual day-to-day operations of banks is about as close as that between weather forecasting and umbrella manufacturing.


I'm confused. Under link [3] you posted, the section entitled "Fractional Reserve Banking Process" states the following:

>The fractional reserve banking process creates money that is inserted into the economy. When you deposit that $2,000, your bank might lend 10% of it to other customers, along with 10% from five other customers' accounts. This creates a loan of $1,000 for the customer needing a loan.

>The bank essentially created $1,000 and lent it to the borrower.

The above explanation was always my understanding. Is this a case of semantics?


> Fractional reserve banking absolutely does not create money[1]. It creates credit exactly the same as if you lend a friend of yours some of your money. You now have an asset (the loan) and your friend has a liability (the debt) and the amount of M1 or M2 money in supply has not changed.

When a bank gives someone a loan M1/M2 increases (unlike in your loan-between-friends example). The increase in "currency in circulation plus deposits" is the very thing that those numbers try to measure.


You assert that the act of individual act of lending creates the money - this is known as the credit creation theory of money; the GP asserts that the central bank creates the money and banks are just moving it around - this is known as the financial intermediation theory of money. That also happens to be the theory that underlies most banking regulation, like the various Basel Accords.

You can look at it either way; or indeed you can take a third view, the fractional reserve theory of money, which suggests that the banking system as a whole creates money in aggregate, but not individual banks.

All of these are theories with their adherents and none has yet been proven right or wrong. The only wrong position is a failure to acknowledge that discussion is still open on this point, or to believe that these are anything other than macroeconomic models.


As far as I can understand seanhunter asserts that bank lending doesn't create money - pointing to the definition of M1/M2 money - and claims that bank lending is not different from me lending you $100 and that the amount of M1 or M2 money in supply doesn't changed in either case.

> The only wrong position is a failure to acknowledge that discussion is still open on this point

Saying that bank lending doesn't increase M1/M2 money supply is wrong. I don't think that discussion is still open on that. It's just how those things are defined. That's the only thing that I asserted.


>As far as I can understand seanhunter asserts that bank lending doesn't create money - pointing to the definition of M1/M2 money - and claims that bank lending is not different from me lending you $100 and that the amount of M1 or M2 money in supply doesn't changed in either case.

Right. That is the basic definition of the financial intermediation theory of money. It's actually a predominant view in the literature: for example, the Diamond-Dybvig model is based on the assumption that banks are not special as intermediaries; and it won the Nobel Prize for its authors in 2022.


I don't know if the financial intermediation theory of money has its own definition of M1 and M2 but the one given by seanhunter is the standard one. If we're using the same definition [are we?] then it either changes or it doesn't.

If money supply is "currency in circulation plus deposits[, etc.]" how does bank lending not increase the "currency in circulation plus deposits[, etc.]" amount?

(Of course lending between friends doesn't: the $100 bill in circulation is the same as before and the friendly IOU is not a deposit nor included in the [, etc.]")


A loan only increases the money supply if it's made without replacing the deposits that were loaned out. The financial intermediation theory is that that doesn't happen: a bank is just a place where money flows come together to find allocations to investments, and that the benefit is that the size mismatches between those in surplus and those who need to borrow can be reconciled, so banks that do this skillfully are economically valuable and make profits.

This is not, prima facie a bad theory, right? Go take a look at JP Morgan's balance sheet. The asset and liability sides of the sheet are basically loans and investments (on the asset side) plus deposits and outstanding debt (on the liability side), and these balance.


Sure one thing can be offset by another.

One may also say that turning up the heater doesn't increase the temperature of a room - because I open the window at the same time.

Anyway, my point was that saying that bank lending is like lending money to a friend and "You now have an asset (the loan) and your friend has a liability (the debt) and the amount of M1 or M2 money in supply has not changed." doesn't make sense.

M1 would change if it was defined as "currency in circulation plus debts between friends" and the "oh, but my friend would sell the debt to the central bank or whatever in the end so there is no change in money supply" argument seems goalpost moving. The original analogy doesn't work and bank lending does increase money supply everything else being equal.


>Still, my point was that saying that bank lending is like lending money to a friend and "You now have an asset (the loan) and your friend has a liability (the debt) and the amount of M1 or M2 money in supply has not changed." doesn't make sense.

On the one hand, you go to a friend and say "hey, can you lend me $100; I'll pay you back $105 in a year?" and your friend agrees.

On the other hand, your friend puts $100 into a 12 month CD, and the bank pays him a 2% interest rate. The bank then turns around and lends $100 to you as a 12 month personal loan with 5% APR.

Can you not see why there are some who would say "it is obviously wrong to suggest money has been created in the second case, but not the first" or indeed "in neither case has money been created"?

By the way, in case it is not obvious: the fact you don't have a compelling rationale to make me believe your description of the world, and I don't have a compelling rationale to convince you of my view of the world is why there are multiple competing theories.

I am not trying to tell you that you're wrong; just that you're not right.


I don't debate any of that!

I'm just claiming that saying that M2 money supply doesn't change in the second case is wrong because it has been defined to measure exactly that. Under the assumption that money has been created in the second case, but not the first - whether we find that obviously wrong or not is irrelevant.

> the fact you don't have a compelling rationale to make me believe your description of the world

I'm only trying to make you believe that the thing that seanhunter wrote is seems incompatible with his own definition of money supply which makes the bank loan situation different from the friend loan situation.


No; M2 has been defined to measure the money stock. You're a believer in the credit theory of money creation, so you obviously think increase in the money stock happens due to lending. If you're a believer in the financial intermediation theory, you don't think that - you think the central bank adds to or removes from the money stock (by controlling short-term interest rates, and therefore the amount of loanable funds), and the banks just move it around.

These are macro theories; they don't tell you anything at all about an individual loan - only aggregate behavior of the entire system.


I'm not a believer in anything.

The only think I've been repeating all along is that in your own example

  On the one hand, you go to a friend and say "hey, can you lend me $100; I'll pay you back $105 in a year?" and your friend agrees.

  On the other hand, your friend puts $100 into a 12 month CD, and the bank pays him a 2% interest rate. The bank then turns around and lends $100 to you as a 12 month personal loan with 5% APR.
M2 goes up in the second case where they end with $100 each (but doesn't in the first case where $100 change hands) because the example doesn't include any mention to a central bank removing from the money stock.

This is the way. It's too bad this can't be pinned to the top. Thanks for laying this out so clearly.


It may be clear but it's not correct.

https://news.ycombinator.com/item?id=35108659


> Fractional reserve banking absolutely does not create money[1]. It creates credit exactly the same as if you lend a friend of yours some of your money.

Yes, it creates credit that people think is money. I suspect that's a big problem with it. People think their money is in the bank. Instead the bank is just extending them some credit that they can pass on to others when they "purchase" something.


Sure, but they don't have unlimited ability to make loans. They are subject to complex balance sheet constraints. And it's cheaper to borrow money from depositors than the Fed.


> Reserve requirements are almost non-existent.

This says otherwise.

https://fred.stlouisfed.org/series/TOTRESNS

Lots more real evidence here https://fred.stlouisfed.org/categories/123


> This says otherwise.

And in Canada, Australia, etc, they are zero:

* https://en.wikipedia.org/wiki/Reserve_requirement#Countries_...


Indeed, in these countries the banks have capital requirements, which they also do in the US, instead.


Those are reserves, not reserve requirements (which are noted as "discontinued" in the first link). To quote this: https://www.federalreserve.gov/monetarypolicy/reservereq.htm

"As announced on March 15, 2020, the Board reduced reserve requirement ratios to zero percent effective March 26, 2020. This action eliminated reserve requirements for all depository institutions."


Read more - the reason is the "ample reserves" regime, where, as I showed, reserves are vastly higher than usual. This is because the Fed has transitioned to a better method to handle what used to be called the reserve requirement. The new (and as the above data demonstrated) is higher actual reserves. This is done via changes in mechanisms, described in the FAQ.

It's not like banks suddenly loan out all the money. The Fed page has a FAQ and you can look up related papers.

From the FAQ: "For many years, reserve requirements played a central role in the implementation of monetary policy by creating a stable demand for reserves. In January 2019, the FOMC announced its intention to implement monetary policy in an ample reserves regime. Reserve requirements do not play a significant role in this operating framework."

People see the change, which is a good move for good reasons, and flip out, just like goldbug times or when CPI get adjusted and so on. A simple way to look at it for a long time there has been a idea battle between what's called endogenous and exogenous money creation (somewhat overview of the debate [1]), and over the past 50 years, most central banking systems have evolved as market needs evolved, to allow banks to lend beyond reserve requirements as long as they soon replenish, which was done via overnight lending windows through central banks. The practical effect is there has not been a "reserve requirement" except in name only for decades. Banks can lend whatever they want, and only have to borrow back to the old requirement which is simply an inefficiency.

This is summed up in one sentence on the Wikipedia article on reserve requirements as "Under this view, reserves therefore impose no constraints, as the deposit multiplier is simply, in the words of Kydland and Prescott (1990), a myth" [2]

Under the new system, the Fed changed other rates to account for this bookkeeping trick, with the same net effect on the banking system, but under simpler and more transparent accounting. They're not idiots.

So, since this was the reality of banking, the FED changed how things are tallied. Just like when CPI changed some terms to handle changes in reality, changes that were well documented, for good reason, people not reading carefully got upset and were sure they got cheated somehow.

So setting the old rate to zero and adopting the new methods for monetary control are the same pattern: terms change, in practice things are actually better off, but people used to the old term are upset and mischaracterize it as some a sign of financial calamity or underhandedness.

It's not. When people dig this out during an unrelated bank issue as some smoking gun it's worth stopping the nonsense in it's tracks in the same vein as COVID or climate or goldbug nonsense.

[1] https://www.sciencedirect.com/science/article/pii/S109094431...

[2] https://en.wikipedia.org/wiki/Reserve_requirement



Not how it works. How it works is that there are a number of constraints and metrics placed on banks that they need to satisfy and prevent them taking risk. At any time it could be any one of these that is the limiting factor on risk. It may not be a measure of "reserves" that is the constraint.


Can you elaborate this. Is it possible for a regular bank to lend money it does not have?


That's not what GP is saying. When reserve requirements are low, banks are essentially no longer forbidden from lending out other customers' deposits.

When you decrease the reserve requirement, you're unlocking reserves that were previously locked up to back your deposits so that they can be loaned out to other customers, who will then promptly either directly or indirectly end up depositing their loaned amount into the banking system again, where the part that remains after the reserve requirement once again gets loaned out, deposited back in, etc.

In effect, because you're making this money go around, you're increasing the total supply of money, because these deposits are the money supply, and there can and often is more than a $1 net increase in overall deposits for each $1 deposited with a bank. Another way to state it is like so: you deposit $1, bank loans out $0.90, someone else deposits that $0.90, their bank loans out $0.81, so on and so forth for $10 of net deposit generation for a $1 initial deposit at 10% reserves held.

Of course, individual banks can elect to lock up more reserves than necessary, which would negate this effect.


> Is it possible for a regular bank to lend money it does not have?

Banks create money from nothing to lend out:

> Therefore, if you borrow £100 from the bank, and it credits your account with the amount, ‘new money’ has been created. It didn’t exist until it was credited to your account.

> This also means as you pay off the loan, the electronic money your bank created is ‘deleted’ – it no longer exists. You haven’t got richer or poorer. You might have less money in your bank account but your debts have gone down too. So essentially, banks create money, not wealth.

* https://www.bankofengland.co.uk/explainers/how-is-money-crea...

> This article explains how the majority of money in the modern economy is created by commercial banks making loans. Money creation in practice differs from some popular misconceptions — banks do not act simply as intermediaries, lending out deposits that savers place with them, and nor do they ‘multiply up’ central bank money to create new loans and deposits. The amount of money created in the economy ultimately depends on the monetary policy of the central bank. In normal times, this is carried out by setting interest rates. The central bank can also affect the amount of money directly through purchasing assets or ‘quantitative easing’.

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

Money is a psychological construct of humans to facilitate trade and the exchange of goods and services. Some societies don't (didn't) even have money/currency: everyone kept a mental 'tally' of who gave or took things, and there were social expectations of giving "gifts" for repayment. Physical tokens (bones, shells, gold, paper, etc) came later.


The posts above are incorrect about how banks work, but your question is spot on. Actually, banks always lend out money they don’t have - they can’t actually “lend deposits”. The risk of what banks do is why banking licenses are hard to get and it’s so regulated.

This primer from the Bank of England is very clear: https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...


They're allowed to borrow from the federal reserve which is allowed to lend an infinite amount.


Banks aren't allowed to borrow infinite amounts of money from the Fed. If they could SVB would still be ticking.


They borrow and repay with interest, so the don't borrow unless they have some use for the money that will return value to the borrower, and some profit to the bank, in order to repay the loan the bank got.


They can borrow easily but it’s still a liability, right? I think people here the borrowing part and don’t know the liability part.


> (Yes, this is a gross over-simplification)

This isn't really an over-simplification as it is wrong. When someone deposits $1m in the bank, the bank's assets increase by $1m in cash, while liabilities also increase by $1m and owners' equity is unaffected. The problem is that for interest-bearing deposit accounts, those liabilities increase over time, which decreases owners' equity in the absence of a sufficiently appreciating asset (such as a good loan or cash flow-generating security). Further, the bank has certain operational costs that must be paid, and investors must make some return or they'll pull capital from the bank (that's a simplification for publicly traded banks like SVB). In practice, it seems that banks need about 3 percentage points above the interest rate they pay on deposits to cover these costs, based on the typical spread between the Prime rate and the Federal Funds rate, though I imagine this necessary yield has a much higher variance for smaller banks.

When interest rates were effectively zero and their deposits increased by a huge amount, SVB decided to buy long-term bonds w/ 1.5% interest to cover the extra liability over time so that assets would grow with liabilities. Then interest rates went up. Cash assets stopped growing, but liabilities remained the same, so they started selling their bonds. Bonds lose value when interest rates increase, so their bonds sold for a loss, decreasing asset values. In the last 48 hours, depositors got spooked. SVB's equity effectively went negative, since asset values decreased below outstanding liabilities. That's typically when the FDIC steps in to liquidate a bank.

For banks, assets must in general be growing faster than liabilities. If the bank experiences a situation where assets are not growing relative to liabilities, they need to have sufficient capitalization (i.e. owners' equity) to weather the storm.


Just to add to this excellent answer, because I don't think it's clear to most people: if the deposit is not created by that bank, the assets the bank takes on are the balancing side of the original deposit (so a loan at another correspondent bank or, if the spending was from the government, an asset at the central bank). Clearing might simplify the net situation, but fundamentally the asset the bank takes on matches the original asset that created the money.


Not an expert, but was having some thoughts.

Let debt be a graph where the nodes are people (with ledgers) and the edges are all of the form "alice rents $x from bob for y% APR". Actions that resolve/relax graph are payments of the form "alice pays bob $z", that lead to all balances being 0. Let the edges decay to null when balance is 0, such that a 'resolved graph' is simply a list of nodes with no edges, meaning 'no one is in debt to anyone'.

From this we can infer:

   1. There is only one logical 'debt graph' in the world since they can (and do) all join.
   2. The people running the graph do not want the graph to die, ever.
   3. The profit of the debt business is proportional to transactions over time,
      which is proportional to edges of the debt graph.
   4. They want to (add, prevent from decay) as many edges as possible.
I somehow feel like I've caught my first, hazy glimpse of something important.


Your model feels trueish to me, but I think the analogy is actually overcomplicating it.

The basic nature of finance is that some people have money they don't immediately need and others have needs they can afford over time but not upfront.

It's basically a layer on top of money in general, which is a way of decentralizing value production. Instead of pairwise trades, money serves as credit for value creation, recognized by 3rd parties.

The people running the graph are everyone with excess money or the ability to acquire it. Basically, everyone but the poor.


Convert “$x” into “something Alice wants”. Alice doesn’t really want $x, she wants a new car, a house, a vacation, a new machine for her factory.

Now “They want to (add, prevent from decay) as many edges as possible” becomes “They want to encourage different parties to lend each other things they want”.


hold up hold up. there's a key distinction between the two categories of things you list in your examples. The first three are items that are consumed to meet Alice's specific desires, with features that match these desires-- a specific car, a specific house, a specific vacation. Alice buys them because she wants to use them.

In a sense Alice also buys physical capital to use them, but she's not using physical capital because of its specific qualities but instead because of the money the products it produces will fetch when sold on the market. Alice wants a vacation because she wants to go to Paris or New Zealand or whatever. Alice wants a machine part because she wants the money it will make her. Two very different categories of thing!

I bring this up not exactly as a critique of your analysis, but instead because none of the discussion in this thread is particularly tethered to the process of production, and it's specifically the disruption to production that makes the SVB collapse troubling.


I think what makes this sound so profound is that the second point sounds like a shadowy, centralised cabal. In reality it's a lot of the participants in the graph, for various reasons.


I also feel like having caught a hazy glimpse of something monumental. Thank you.

A few thoughts to add:

1. Should individuals inherit edge weight from their employers/governments? What should this inheritance be like? If I have no debt but have very little in savings, and if my company has debt to some other company which they have to default on, leading to my getting laid off, I can still be affected by debt.

2. I do not think it is accurate to think if "the people running the graph". I think it's more accurate to state local laws such as: any node which has more outgoing weight than incoming weight (i.e. net lenders) wants to prevent edges from decaying as much as possible while simultaneously making their borrowers more financially stable.

3. Would it be possible to quantify each vertex's credit-worthiness from just the labels you mentioned on edges? Would we need to add any other weights to the edges? e.g. the lender's estimate of the borrower's credit-worthiness?


Thanks for that, but after thinking about it a bit I don't think I have much to offer. Perhaps if we were in a pub and I could understand where you're coming from. The thing about models is that they can be used in a lot of different ways, and can be pushed in many more ways. The model seems so trivial that it must be part of a 1000 other models if it is correct, or it must be well-known why this model fails. I imagine the quants rolled their eyes and didn't even bother responding, like how physics people do when laymen riff on quantum mechanics. (My ego would love it if this is novel, but my brain knows better.)


I like the graph idea, but I might be missing some of what your saying... Although I think that it should be more than just people as nodes. Anything that can participate in the economy via signing contracts woukd be my first approximation. So, businesses, cities/states/nations, etc...

Also, I think it's reasonable for individuals to inherit edge weights from their governments, as a way to model taxes, national debt, etc.


> 2. The people running the graph do not want the graph to die, ever.

The problem with formulating it like this is it makes readers think that the set of people is a small-ish set of globalists or capitalists (or insert conspiratorial "others" as appropriate). But the set is far from small: basically anyone who ever wants to acquire debt for any reason (most common reasons include "attend college", "buy a car", or "buy a house"), or anyone who wants to profit from lending money to people, which is a fancy way of saying anyone who wants to invest money (such as buying stocks or bonds).


If we're modeling the flow of value (in the more abstract), then holders of most currencies on the planet have real value constantly flowing away from them... by virtue of inflation.

Or in a larger macro sense, positive central bank interest rates create a requirement for everyone to acquire some debt in order to counterbalance the drag.


Yep, this is indeed what is going on, and relates to concepts like "capitalism motivating nationalism, imperialism and colonialism" - it's the large scale expansion of the debt graph. It's something you can pull off profitably temporarily by coordinating your nation towards a stronger state bureaucracy, aggressive trade policies, industrialization, and destabilization or dehumanization of rivals. When we speak of "this is good for the nation", what's really at stake, materially speaking, is this kind of Machiavellian machination of debt, tempered only by whether it can be justified in terms of national myth, dominant ideology, party politics, etc.

It produces a grand, generational historical narrative, but it's also a form of pyramid scheme and guaranteed to run out of runway at some point when the graph, rather than expanding, finds ways to get by with different trade flows that ignore the center - and each time that happens, you get a massive economic crisis, elites vying for power and drumming up scapegoats to avoid heat, but also waves of material change(different lifestyles and work arrangements).

So there is downside to debt in that it can reinforce hierarchies, but also upsides in that the network itself is acting to transfer useful information about material needs. All things that, having becoming so much more digitally connected in the last 30 years, we can probably revise again to become more abstracted.


Proposition 2 is flawed: some people DO want the graph to die.


Yes, it strikes me as perverse that if your business is debt, then a debt-free world is a very bad world. It's perverse in the same way it is perverse for a doctor to be sad when fewer people get sick. So, I empathize. Be that as it may, I don't see fewer people getting sick or wanting debt, so I think the point is somewhat moot.

It is troubling that the debt industry will attempt to increase the total debt in the world however it can. This is like a doctor attempting to increase the total sickness in the world - which is evil (not a word I use lightly, btw). But to say that either debt or medical industries should 'die' because perverse incentives exist seems wrong to me.


To a large extent it reflects liquidity demands. When governments are falling over themselves to profess "fiscal responsibility" the liquidity demands are met through private banking.


There's a concept in Anthropology/Primitive Economics called "Gift Economy" where participants in a community gifted each other help, goods, food, and rely on being gifted back similar things in higher/lower proportions sometimes in the near future.

If I'm not mistaken, Debt: the First 5000 Years and the Dawn of Everything books examined this.

Your graph is a formalization of such relationship, no?


2. The people running the graph do not want the graph to die but want the edges to decay.

Thanks for this, I am now also having some thoughts.


This system will settle on a solution if there is no net total creation of money in the world. As long as new money is allowed to enter this system, edges will keep appearing to protect it.


1. This doesn't actually mean anything. It's true by your definition.

2. There is noone running the graph

3. This is true of all econonomic activity in a free market economic model. The assumption underlying market capitalism is that on net rational actors will only perform transactions which add value to them. Therefore the profit of the system is proportional to transactions over time which is proportional to the edges of the "financial transaction" graph. Debt is not required for this observation to be true - it's intrinsic to the model or the transactions won't occur.

4. There is no "they" other than "all the participants in the financial system". On net everyone participating in the financial system is doing so for their own benefit and therefore are invested in keeping the system running.


This is actually a nascent PhD-level thesis. Follow this thread.


What makes you say that? It seems to be a restructuring of a supply/demand curve along a lattice. (Ironic, given your name lol?) https://en.m.wikipedia.org/wiki/Total_order


This isn't an over simplification it's just wrong. The bank does have more assets (the 1 million in cash) that balances the liability (the 1 million they owe to their customer).


Yes, but the problem is the 1 million is unlikely to be in cash. Here, it is in MBS issued at low interest rates and thus practically ~$500k if the security is sold right now.

Banks have a serious problem in their hands as they have to figure out a way to keep buying assets that they have to buy by law, while the Fed is going to keep increasing the interest rates via selling their MBS portfolio at a rate that makes "yesterday's treasury or MBS" the loser.

If action is not taken we haven't heard the end of this.


The $1 million deposit creates a $1 million cash asset and a $1 million account liability.

The value being in MBS now is a result of a separate transaction (well, a lot of them, probably) where (in aggregate) the bank takes a bunch of money from its cash assets and trades them for an (initially) equal value in MBS assets.

It becomes a problem when the MBS assets lose value, and the bank needs cash to cover withdrawals.


This is right, though I think we should be clear that the MBSs lost value because they are normally like bonds, and bond prices decrease when yields increase.

MBSs tend to remind people of '08 when MBSs lost value because too many had a bunch of garbage loans which defaulted, causing MBSs' face values to decrease. In this case, the fact that they were MBSs is mostly a coincidence. If SVB had invested in Treasuries, they'd be in the same world of hurt that they're in today.


> If SVB had invested in Treasuries, they'd be in the same world of hurt that they're in today.

Not necessarily, if they were investing/laddering in short term Treasury Bills (< 52 weeks) instead of longer dated bonds things might have panned out differently.

”97% of these MBS were 10+ year duration, with a weighted average yield of 1.56%.”

Treasury Bills from 6 months ago had yields of 3.37%.


What?

No, it's cash.

If I deposit $1m in cash at the bank, the bank suddenly has an extra $1m liability and an extra $1m cash asset.


And you think that the bank just keeps sitting on that cash until you withdraw it again, paying you interest in the meantime?


If they are sitting on excess cash, they can lower interest rates on deposits.

The point is that when banks receive deposits, in the short term, their assets/liabilities doesn’t change.


Sure, but you don't get a useful model of a bank by only looking at the short term like this.

> The point is that when banks receive deposits, in the short term, their assets/liabilities doesn’t change.

They do change: Cash is a (risk-free, modulo safe storage) asset, the deposit is a liability. This has consequences for all kinds of metrics vital to the running of a bank.


Correct, although if they don't do something to make money with the cash they take in as deposits, they won't have sufficient income to cover their expenses, let alone make a profit.


The effecf of the deposit is separate from the effects of decisions the bank makes after the deposit.

The deposit creates a cash asset and an equal liability for the increases account balance.

Other transactions have other effects.


It’s cash, until the bank converts it to something else. For example a loan.

Yes, it is initially cash. But if the bank wants to make any money, it needs to convert that cash into some other vehicle for generating interest.


One way to generate interest is to simply park it at the Fed. Check out The Narrow Bank for the story of an upstart that wanted to make that their whole business model.

https://johnhcochrane.blogspot.com/2018/09/fed-nixes-narrow-...


Thank you for this! Very interesting.


You're right, the full million isn't a liability. But anyone depositing 1M isn't putting it in a non-interest bearing account. So now you are generating liability every second you sit on the cash. Which means you need to "invest" it.


No one is showing up at a bank and depositing $1M of physical cash (except maybe Mexican drug cartels).

That $1M you're depositing is presumably the liability of another bank.


Doesn't matter if it's physical cash or an electronic ledger since both the electronic ledger and the physical green pieces of paper are liabilities of the banking system, specifically for US dollars they're both liabilities of the Federal Reserve.


> I don't see how those are liabilities of the Federal Reserve, but maybe I'm missing something?

To my sibling: Missing is the fact that US dollars are bank notes issued by the Federal Reserve. Issuing a note creates a liability.

https://www.federalreserve.gov/monetarypolicy/bst_frliabilit...


> specifically for US dollars they're both liabilities of the Federal Reserve.

I don't think that's right. There are banks outside the US that hold US dollars that the Federal Reserve has no jurisdiction over. They create more currency through fractional reserve lending for their local currency as well as any dollar denominated loans they make. I don't see how those are liabilities of the Federal Reserve, but maybe I'm missing something?


Dollars are bank notes “promises to repay”. It used to be many different bank issues notes payable in gold, now it’s one bank - the Fed, and it’s no longer gold.

It doesnt matter if the US dollars is held outside the US. The dollar represents a promise by the fed to pay, and thus is a liability to the fed.

Of course the fed has the power to create and destroy dollars.


> Of course the fed has the power to create and destroy dollars.

So to banks outside the US. That's how fractional reserve banking works. If I go to the UK and deposit $1m USD, and they loan out $900k of that to you, that UK based bank has created $900k.


Yup, that’s quite the insight. (Although I don’t understand why you write banks are forced by law to buy.)


A $1M deposit adds to both assets and liabilities.

In double entry accounting, you mark any changes with both a debit and a credit. This allows you to see not only why one account changes (a single entry), but also the cause of that change (the second entry).


I think programmers find double-entry bookkeeping counterintuitive. It feels error-prone. In programming, you keep a single source of truth. Any time you copy the same data to two different places, one of them is always wrong.

Double-entry bookkeeping makes sense once you understand the the invariants you have to keep, and why you need to track 5 different types of books. Some of those accounts work in opposite ways, such that credit to one is a debit to another.

It all works out and is essential for "debugging" problems (when money appears to go missing -- or worse, materializes and you don't know why). But there's some counterintuitive language and it'll mess you up until you accept it.


The name "double-entry" is a little bit misleading. The "double" part makes it sound like you are duplicating work, but that's not really what's happening.

If you were accounting with pen and paper, you would write the transaction amount twice — positive in one column and negative in another. This maintains the invariant that money cannot be created or destroyed.

In terms of an abstract data model, though, each transaction is best thought of as a flow, or a weighted arrow. It has one amount and two ends: a source and a destination. The "double" in "double-entry" really just means that the arrow has two ends. Obviously every arrow has to have a head and a tail — it doesn't make sense for there to be no source or no destination.

The credit vs. debit thing confuses people a lot, and in my opinion is a red herring. If I could wave a magic wand, I would delete the words "credit" and "debit" from accounting because they are hopelessly inconsistent.

All you have to do is think of the conservation of money like the conservation of mass — when it moves, it leaves one place and arrives at another. The moment I stopped using the words "credit" and "debit", my understanding of accounting went from "I have no idea what I'm doing" to "Everything is intuitively obvious."


Yeah I agree. I aced my MBA accounting class but it makes a hell of a lot more sense to conceptualize everything as “stocks and flows”.

Also, I would say transactions are actually “multi-arrows” since there can be multiple sources and/or multiple destinations. A Bitcoin transaction captures this structure more or less perfectly.


Yeah, I just treat credit and debit as opaque blobs. They're words, I don't really know what they mean, but I know one means "side A" and one means "side B" ;)


> This maintains the invariant that money cannot be created or destroyed.

And if you wonder, if money cannot be created or destroyed, how does a monetary system handle the population tripling over the past 50 years?

The answer is national debt. The US national debt is just an artifact of this double entry bookkeeping. In order for there to be +money in the economy for the ever-growing number of citizens to do commerce with, the treasury incurs on itself -money. That's national debt.

The corollary then, if the US ever pays back all of its national debt, the economy will crash because there just aren't cash available for people to do commerce with. A +90trillion on the government balance sheet equals a -90trillion on the private sector balance sheet.


> And if you wonder, if money cannot be created or destroyed, how does a monetary system handle the population tripling over the past 50 years?

There is ONE special actor (in the US) that can create money out of the thin air: the Federal Reserve.

It alone can "buy" securities by crediting banks with money created out of nothing. Banks simply see a transaction with a dollar amount on their accounts within the Federal Reserve, and that's it. The money just appears.

The invariant "money can't be created or destroyed" holds for everybody else, including individual banks.

> The answer is national debt.

That's completely incorrect. The Federal Reserve can arbitrarily create (or destroy) money even if the national debt goes away entirely.


Most of the Federal Reserve's balance sheet is Treasury securities. If the Treasury stops issuing those securities, then the Federal Reserve don't have anything to buy.

Saying that the Fed can arbitrarily create and destroy money misses the point, that they then have to buy SOMETHING with those money, and there are strict rules about what they can buy, and it just happens that the Treasury controls the supply of the primary asset class that the Fed is allowed to buy. Once you clear out all the dance and ceremony, you arrive at the conclusion that the Treasury issues new money by issuing new securities.


> Most of the Federal Reserve's balance sheet is Treasury securities.

That's true, simply because they're the most convenient way to manipulate the monetary supply. Not much else is readily available in the volumes needed.

But they are not _essential_.

> Saying that the Fed can arbitrarily create and destroy money misses the point, that they then have to buy SOMETHING with those money, and there are strict rules about what they can buy

Sure, the invariant: "money goes out, asset goes in" holds.

But they can buy quite a lot of different securities if needed. E.g. the Fed directly bought about $3T of MBS: https://www.newyorkfed.org/markets/programs-archive/large-sc...


Not really. Money is created but it is created by the fractional reserve baking model where banks borrow from people who are willing to lend for short periods of time (depositors) and lend to those who are looking to borrow in the long term. The fact that you tend to deposit in checking accounts is what creates the money.


The private banking system can only multiply off of the monetary base up to a certain multiplier, it cannot expand that base. Only the government treasury can expand that monetary base.

If the private banks have conjured up 100x the monetary base in 1970, then when the population doubles by 2020, the banks cannot multiply further up to 200x. The treasury has to expand the base while the banks keep their multiplier fairly constant.


Double entry bookkeeping IS error prone - you have 2 opportunities to make each mistake.

However single entry bookkeeping is FRAUD prone. Just change one number and your theft is hard to track down!

That is why the adoption of double entry bookkeeping was critical for allowing commercial institutions to outgrow a size where owners could individually trust all who were working for them with access to money.


Yes, and once you get the programmer to realize this, it all makes sense.

The ledgers are not the source of truth. No: this is a journaled filesystem; the transaction log is the ultimate source of truth. When you detect faults you recover from the journal, if something isn't completely written into the journal then it does not exist. The zero-sum property exists on every individual transaction and each transaction has an ID and the ledgers point at these IDs for auditing purposes.

So why have the zero sum property? Well, for one thing, it creates a uniform access model, I can't just credit my account, I have to debit someone else's account and they can have rules that might prohibit me from doing so. By carefully setting up these accounts you can also do what's sometimes called "behavior anomaly detection."

So for instance normal financial cards kind of work by opening up your entire wallet to a cash register and asking the cash register to pull out exactly the amount that you owe, understandably this might not be desirable if you are tracking some in-game-gold transactions in an internet game. You can get as elaborate as you want but think for example of a quick-consistency-check rule saying that “accounts starting 4xxx (player balances) never transfer directly to each other, instead users are expected to put the exact sum of money plus a little 5% padding into one of their 5xxx accounts and then give someone else a token permitting them to withdraw from that account." Stuff like that. A developer tries to code something up that doesn't go through this process and runs into errors in testing and has to conform their code’s behavior to the less risky process so that nobody is ever opening their entire wallet to a griefer.


The zero sum property also makes it really easy to just add up along (any subset of) each column and find most errors. Not so important now when everything's stored in a computerized database anyway, but a lot more relevant to actual books with actual writing in them.


That’s an oversimplification. Saying that it’s error prone because you have to write in two places is akin to saying that using a checksum is error prone: you have to write now the vale AND the checksum! More opportunities to make an error!

Doble entry accounting is very similar to a checksum in that sense. It provides error detection. If you make a mistake in one of those two places, you will know immediately. As opposed to single entry, where you can carry that error indefinitely until somehow you catch it.


> That is why the adoption of double entry bookkeeping was critical for allowing commercial institutions to outgrow a size where owners could individually trust all who were working for them with access to money.

Wikipedia tells us that double-entry bookkeeping is first attested in the late 13th century.

But commercial institutions larger than the personal-trust threshold are attested as far back as written history goes.


> commercial institutions larger than the personal-trust threshold are attested as far back as written history goes

[citation needed]? I don't necessarily doubt you but I'd like to see an example. The only thing I can really think of would be governments, but that was definitely personal trust (mixed with a little bit of threat of summary execution if the king didn't like you).


I guess I'm confused. If you're willing to consider governments (makes sense to me!), we are informed that the Achaemenid Empire (roughly 2 million square miles, mid-6th-century to mid-4th-century BC) was divided into 20 "tribute" districts: https://en.wikipedia.org/wiki/Taxation_districts_of_the_Acha...

The smallest amount of tribute listed there is more than five and a half tons of silver. If we make the assumption that each of these districts was handled by a different person, we already have a staff of 20 people plus the king.

(Not to mention, the large majority of Achaemenid taxation was not collected in this form. It was collected in kind and stored in an empire-wide system of distributed warehouses. I assume metal tribute was centralized. If you think the warehouse administrators - who inventory the goods and are responsible for making payments out of them - are "responsible for money", you're probably adding at least a couple dozen more people.)

By contrast, https://www.theofficialboard.com/org-chart/exxonmobil suggests that the top two levels (including the board) of Exxon total 26 people, or (excluding the board as a level, but still including the CEO) 24 people. The CEO has 13 direct reports.

So if the argument is that ancient kingdoms were below the personal trust threshold because the king had a low - and therefore manageably trustworthy - number of direct reports, it appears to be the case that our largest corporations today are also comfortably below the personal trust threshold, so there was never any need to exceed the threshold and in fact we never have.

(This isn't quite an apples-to-apples comparison; maybe there are 20 tribute administrators who report to a high overseer of tribute who reports to the king. Maybe each tribute administrator is one of five reporting to a low overseer of tribute, the four low overseers report to a high overseer, and he reports to the king. I don't know. How many employees do you think Exxon has who are directly responsible for submitting revenue totals in excess of three million dollars?)

On the other hand, if we think Exxon is beyond the personal trust threshold by virtue of its geographic extent (huge!), or its total number of employees (huge!), or its total number of employees touching money with the potential opportunity to steal that money (still huge!)... I think we have to say the same thing about ancient palaces, and really about ancient temples. A temple might "only" employ a few hundred people, but that's more people than you can personally trust. An important temple employed a few thousand people.


Governments operated differently than businesses.

In a business I hand you my money and trust that you will hand me back my profits.

A ancient government put someone in charge and demanded tribute. Failure to produce tribute resulted in an army showing up to collect tribute. It was so common as to be expected that the administrator, called a satrap, would collect excess tribute and live a lavish lifestyle. The king sent spies to root out the worst of the abuse, but it was an uphill battle.

Governments can be run this way because efficiency is not particularly essential to their profitability. But efficiency is required for commercial enterprises. It is not enough to collect money and let your subordinates enrich themselves. You want to track all the money and not let your subordinates cheat the enterprise to their personal profit.


The claim above is that empires are below the threshold where everything runs on personal trust, not that governmental administration is fundamentally different from other kinds of administration.

> But efficiency is required for commercial enterprises. It is not enough to collect money and let your subordinates enrich themselves. You want to track all the money and not let your subordinates cheat the enterprise to their personal profit.

The first two sentences are obviously false; you want to track all the money, but you have no particular need to do so as long as some of the money is making its way to you.


So, in essence, the single entry bookkeeping will be sound if the change is not allowed?


Single entry bookkeeping isn't really a thing and isn't really keeping books. It's just an account register or account statement with no totals or anything to check against.

Double entry basically just means that for every transaction on your books, you are showing where the money/asset/liability came from and/or where it's going. If you enter that you paid $100 for the AWS bill, you need to indicate from what account that money came from. So you would debit AWS Expense and you would credit your Wells Fargo Checking Account. With modern software, this basically just amounts to entering the expense and specifying which account it came from.


In my mind it would be more like only a total - 'single entry' is how most people manage their wallets: '£5 change, put that in there, oh I have £15 now'. No record of where anything came from or where anything's gone.


This sounds like "0 code". No code is the best code because it never breaks, has 0 serve space requirements, will never need to be reformatted, etc.


> I think programmers find double-entry bookkeeping counterintuitive. It feels error-prone. In programming, you keep a single source of truth. Any time you copy the same data to two different places, one of them is always wrong.

Double-entry bookkeeping is creating of an audit-trail/error-evident data store; which is the purpose of the apparent duplication. But it can be viewed as a view of dataset reflecting a single source of truth, that documents value flows; every movement of value has a source and a sink, and the amount that moves from the source must equal the amount that moves to the sink. Losing the redundancy that allows error-checking of records, you could view each accounting transaction as a triple of (source, sink, amount).


I think if a programmer were designing accounting, they'd put the flows in the database, and build views out of that. They might materialize the views for performance reasons, but they'd treat those views as suspect and the first response to any bug would be "blow away the accounts and recreate them".

You couldn't do bookkeping with actual books that way, and historically this way makes sense. Nor is it likely that the accountants are going to rethink their field from the ground up for the convenience of programmers.


> I think if a programmer were designing accounting, they'd put the flows in the database, and build views out of that. They might materialize the views for performance reasons, but they'd treat those views as suspect and the first response to any bug would be "blow away the accounts and recreate them".

That is actually how it works, if you don't have "flows" you don't have accounting postings and if you have to fix an error you redo the postings from the flows. In this sense I would say that the "flows" are the source of truth.


Right. The only actual raw data is transactions. Account balances are just filtered rollups. Double entry bookkeeping is just how you manually accumulate a set of account balances from transaction data, and a set of conventions for consistently labeling and grouping the transaction sources and destinations.


I mean, this is kind of how accounting does work. To a basic approximation a reconciliation is a line by line review of each of the flows, and an audit is "blow away the accounts and recreate them" and both happen all the time.


The other thing you get into as scenarios get complex is that there's a tendency to ask questions like "Why are you recording it this way rather than that way? The second way seems more logical to me."

The answer usually being some variation of "Because that's the way the Financial Accounting Standards Board says to do it."


I've been running my own finances with double entry bookkeeping (via beancount) for about 4 years now, and it's ended up being super useful. As long as I'm 90% sure I got all the income, expenses, investment buys and sells, and transfers between accounts, I can add "balance" statements for a date and easily work backwards to find if I missed anything.


I have a friend that's an accountant that tried to explain how it worked. I thought it was more or less being responsible for counting money, but I think my eyes glazed over a bit trying to understand it.

But to be fair he did pretty much the same when I tried to explain programming.

Always important to step outside your software bubble once and a while.


I'm trying to understand how to read older 10-Ks and apply the capitalization of operating leases to those reports (converting property that is leased into capital to get an accurate estimate of the capital employed at the business) and with my background in IT and physics it is still a bit hard to follow (what I'm slow at is language, and really this is just learning a foreign language).


Double-entry bookkeeping is essentially the law of conservation of energy applied to balance sheets. It's much more deep as it was invented some 5 centuries earlier than programming.


Conservation of money in that case. Money doesn’t just disappear. What goes away somewhere has to appear somewhere else.


exactly


If by 'error' you mean the amount of a transaction then yes you enter it twice but that should not really increase errors.

Now, on the other hand, it is especially designed to account correctly for that amount in order to keep your accounts correct and balanced. It is be much easier to lose track of things with single-entry accounting.


I think programmers find double-entry bookkeeping counterintuitive. It feels error-prone. In programming, you keep a single source of truth. Any time you copy the same data to two different places, one of them is always wrong.

In essence you are just recording transactions with source and destination accounts and the amount. And at least in principle you could just log all transactions in this way without any redundant information. If you insist on making the change to each account more obvious, then you will have to record the amount twice, once for each account, but if you are doing the accounting with a computer, then I would guess that you always just record the transaction and only show the amount twice with different signs for the two involved accounts. Or does bookkeeping for some weird reasons really record each amount twice?


> Or does bookkeeping for some weird reasons really record each amount twice?

There is a surprisingly common reason, but it isn't weird. Consider a paycheck. You have income, but then you split that up. Some pays for insurance premiums, some is withheld for taxes, some may be contributed to a retirement account, and some is deposited into a bank account. Earnings is a credit to an income account, but everything else is a debit to some other account. The total debits equal the total credits, but each credit and debit must necessarily be a separate entry, because they all affect different accounts.


> It all works out and is essential for "debugging" problems (when money appears to go missing -- or worse, materializes and you don't know why).

That's equally true in programming. The problem is that it only helps you debug problems that it created or participated in...

The difference in accounting is that it's not "two entries for the same thing". It's a credit and debit entry. It's a source and a target entry. It's two different entries for two different things, even though some of the details (like the transaction that they actually relate to) will be shared.

(And they're treated as separate "rows", rather than as additional columns on a single row, because it's a many-to-many relationship. Very database-designy!)


Programmers like a single source of truth — until duplicating data is necessary for reliability (redundancy) or performance (replication).

There is a reason why planes use 3+ CPUs running the same code, or why Google runs more than just one cluster of search.


You'd think that the solution to that confusion would be database transactions (commit everything or nothing), but there are exceptions. In ACH, sometimes you have to send out a lone credit or debit with no matching entry. It feels very wrong, but it's necessary because it balances out some other entry in another file (that you might not have access to.)

OTOH, once you have double-entry bookkeeping in your brain, you won't want to go back. I now feel that, from the perspective of a business or consumer, money is never created or destroyed, it only moves between accounts.


The reason for the invention of double entry bookkeeping was to detect arithmetic errors in summing the ledgers.


They have both $1m more assets and $1m liabilities.

But that does not reflect risk.

For instance, now they take those $1m in cash and use them to make risky loans or investments. At face value the balance sheet is the same because they still have $1m in asset... except that the risk that this asset turns into eff all has significantly increased.


yeah, like SBF's "balance sheet". https://www.ft.com/content/0c2a55b6-d34c-4685-8a8d-3c9628f1f...

Theoretically they had assets, but most of them were just internal magic beans.


Too much money is only a problem if you are greedy for returns, like all the investors who lost money when yields were unsustainably low the last few years. They could have deposited it with the Fed and have been totally fine.

Individual investors don't even have that option and also have inflation to deal with. Banks don't.


6 month T-bills were yielding 0.7% last year at this time. Inflation for 2022 was over 6%.

It’s not that people are greedily chasing returns, it’s that they don’t want to lose significant chunks of their capital due to loss of value.


No, they were greedy. They were going to lose it one way or another, either through inflation or through asset depreciation. There was no way around it. Going to a riskier asset doesn't help because all assets have the same interest-rate risk embedded. They just took on more risk in addition to that.

Real rates were negative for 10+ years (and still are on the short end!). That's everyone paying for trillions of dollars of wars, speculation and bad investments. The bill has to be paid.

But as mentioned, none of this applies to intermediaries like banks. They aren't forced to take on any risk.


I think coredog64 is speaking more generally. The bank in the article was foolish. Other banks can still encounter real problems with too much cash on hand... granted they don't have to go hog wild on MBS or anything like that.


Yes, why did they invest in MBS, with negative convexity, at historically low rates!


> Real rates were negative for 10+ years (and still are on the short end!). That's everyone paying for trillions of dollars of wars, speculation and bad investments. The bill has to be paid.

They sure were, and fortunately humanity took that free money and invested it in sustainable energy technology so that at least we got a bunch of infrastructure out of it.

Oh actually wait we just had austerity for some reason.


It's not theirs, it's the depositors. And it's the depositors choice to leave it in the bank.


I think there is a limit. The bank costs money to run, customers expect interest (even if little).

That money isn’t doing anything to cover those expenses if some of it isn’t “working” to produce returns.


Banks don't pay the Fed rate on deposits so they already have that to pay expenses. In fact more deposits = more money on the interest differential but expenses don't scale with money.


A deposit of $x adds both $x in assets (cash) and $x in liabilities (account balances).


One way to control it is by lowering interests on accounts to make it less attractive deposit money


Which is why there is a ~4% spread on savings/CD interest rates at banks right now.

Edit: maybe I wasn't clear and should have said "between" banks. Synchrony bank will give you 4% in a _savings_ account (https://www.synchronybank.com/banking/high-yield-savings/?UI...) while my local credit union and many big banks (ex: bank of America are at 0.01%) are still effectively 0%

In the past (<2008), cash accounts (savings/cds/etc) I remember there being a %1 or so between banks, but these days its huge.


Synchrony offers such high yields on savings because their big product is consumer credit cards, and even at 4% APY on savings accounts the spread to a 25% APR credit product, even factoring in risk, makes it extremely profitable to do so.

That's really the big reason why savings rates vary so much between banks, if you have a bunch of consumer credit lines then high savings rates make sense, if you have a bunch of lower fixed-rate collateralized loans then you can't get away with lowering your spread so much.


It's not nearly that simple, though. You are ignoring capital requirements. For things like Treasuries, you don't need any capital % against it. ($100 in deposit -> $100 in Treasuries, if you borrow at 1 and lend at 4 you made a 3% spread on $100). But for risky things that might yield much more, you might need to reserve far more, meaning you need to fund part of it with equity. ($100 in deposit but with a 50% capital requirement means you can -> buy $100 of high-yield stuff, but you need to use $50 of your equity to do so)

There's other epicycles to this too. For example, a bank can just "buy" deposits through a correspondent bank (a bank-for-banks). So the direct connection between your kind of borrowing and kind of lending, particularly when its consumer / credit card stuff, can be tenuous. In SVB's case, though, all of the stuff happened in a tight echo chamber ecosystem -- SVB got deposits from tech cos, who got money from VCs, who were working off of capital call LOCs from SVB. So your observation is germane.


Yeah, I oversimplified it to a large degree because getting to deep into the nightmare that lies beneath the banking system can make ones head explode. It's accurate enough for a layperson, if nothing else.


Not exactly. The banks do not want less deposits (ask SVB how bad that can be!) but they try to get away with paying as little as possible and keep the spread. (You wouldn't say that the main objective of companies when they raise prices is to have less customers, would you?)


If they wanted more deposits, why would they pay as little as possible to 'keep the spread'? No bank is anywhere near monopoly; there are tons more deposits to get (if you want them, and can make money safely on a spread).


For the same reason that many businesses do not want "more customers" above anything else - to the point of selling at cost or at a loss - and if they increase prices is not necessarily because they want "less customers".


Some banks actually do want less deposits. They do things like increasing fees to chase them away. But it takes some real discipline and investor management to pull it off.


If a bank is noticing they are getting more deposits then they can safely make returns for (like SVB in 2020), can banks "simply" return deposits to customers and/or decline new customers/deposits? Is there a precedent to doing this? Or does everyone decide it's better to take risky bets to grow with the increasing deposits and we'll keep seeing runs when those bets go south?


> When someone deposits $1m at a bank, the bank doesn't have $1m more assets, it has $1m more liabilities.

Not quite. It actually has both.

Yes, it owes $1m to the depositor, but it also has $1m in cash now, at least for a while. Until it does something with it, like loaning it out or investing it in some debt instrument. The cash is an asset, as is the loan or debt instrument it exchanges the cash for.


Running a bank sounds like a huge PITA.


This is correct. The only reason people still do it is because it makes them boatloads of money, not because it is fun in any way.


I work in a bank and it's kind of fun.


Working in a bank and running a bank are slightly different I would assume


Totally. Working in a bank exposes you to the fun parts of running a bank, without the stress of it being your bank.


Let's hope (for all our sake) your work continues for a loooong time, fun and not-so-fun.

Until we all can self-bank!


That's why it pays.


Yes, bank deposits are a liability for the bank, which is why a bank won’t increase the amount you have in your account if you ask nicely. You have to pay them. If someone’s bank account went up by $1 million, the bank better have gotten $1 million more in assets from them somehow, or something weird is going on.

(Often, the asset comes from the bank making a loan. You can pay later.)


I think the additional liability from that event would be much less because it would be computed from the interest payments that the bank eventually owes the depositors.

So given an interest per annum of 1% (for example), a $1m deposit would add $1m liquid, non-earning assets while also adding $1.01m to their liabilities. So, effectively, they’d have $10,000 in unsecured liabilities.

To counter that imbalance (as well as protect the cash from the effects of inflation), they’d have to put some of that cash to work via various investments.


They have both 1m of assets more than before, _and_ 1m of liabilities.


So true.

Money for banks is what raw material is for manufacturing. What they are doing is risk management (analyzing risk, packaging risk, selling it, buying it).


I think it should be both: The $1m that was deposited is an asset, the customer's ability to withdraw $1m is a liability.


No, the $1M deposit is an asset and liability (which offset each other). SVB then invested those assets in bonds/securities/T-bills but then had to sell them at a 20% loss because they needed funds for liquidity—-so now they don’t have as many assets as they do liabilities and hence are taken over by FDIC.


And to make matters more confusing, deposits are called assets on the balance sheet!


You may feel relieved - and less confused - if I tell you that deposits are liabilities in the balance sheet.


That's not confusing, it's just wrong.

On my balance sheet, money I deposit is an asset: The bank owes me.

On the bank's balance sheet, money I deposit is a liability: The bank owes me.


Your deposit creates an equal asset and liability on the bank's balance sheet remember assets = liability + equity must balance.


Deposits are liabilities to the bank's customers. The cash the bank has, regardless of how it got it, is an asset. The cash may result from a customer making a deposit, but its still just cash. Deposits and cash are separate items on the balance sheet. Here is an example from the Fed:

https://www.federalreserve.gov/releases/h8/current/

Note that cash is a single line item. The banks could get that from other sources, such as loan repayments.


The transfers are settled in central bank reserves, so the net change of assets and liabilities should actually be zero for deposits coming in.


It has +1mm cash (asset) and +1mm liability to depositor


Actually if you think about the balance sheet for banks "too much money" is a problem for them because deposits are liabilities not assets. The money belongs to customers not the bank. The bank must pay back the customers on demand.

That said, I think historically many banks can easily avoid the "too much money" problem by setting the interest they pay on deposits to be low, and maybe even negative.


Agreed. Also too much money is hardly an automatic existential crisis. They can do what I described, get bought, merge, etc.

Of course if you dump it all into unwise investments, that’s a problem.


Why is "too much money" a thing? If a bank only wants $100 million in deposits, but customers deposit $150 million, why can't the bank set the extra $50 million to the side and pretend like it doesn't exist until customers want to withdraw it?


Because now the bank has to pay interest to depositors of $150M instead of $100M, which means that they'll pay a lower, less competitive rate. So, in order to keep customers, banks are incentivized to lend out any and all spare cash for whatever yield that they can get, in order to give attractive rates to depositors. Losing customers though shouldn't really be a problem for the bank, after all, those customers did deposit "too much" money - once enough have left to seek higher yields elsewhere, there will be less cash on the sidelines, and so higher yields for the remaining customers. I suppose if your whole philosophy is "growth at any cost", and you're measuring growth not just by AUM but also by number of customers, you get excess risk taking and yield chasing.


So the conversation goes:

A: We have too many depositors! We are not getting enough yield to pay interest without taking on risk.

B: What if we reduced the interest we pay on deposits?

A: Then we'd stop getting new depositors! Our only option is to take on risk.

You are right, this feels like a very unsympathetic problem to have. If you are a regulated bank you need to act like a grown-up and understand that overworking the soil and underworking the soil will both give you a bad yield in the end.


At the SVB bought the securities in question, interest rates were 0 across nearly all savings products. SVB was trying to maximize profits for itself and shareholders. This isn't about attracting customers with high-yield accounts.


... though it may well be relevant that they might have felt great pressure to /retain/ customer deposits in a world of competitive yields.


How bad can it be when most banks are paying 0.2% on savings accounts and nearly 0% on checking?


I feel most banks are coasting off the inertia from longtime and/or unsophisticated customers.

Up until I switched to a neobank late last year to get some actual yield on my savings, I'd been using the same brick-and-mortar checking and savings account I opened in high school.


Apparently your average person is much more likely to move homes than move banks. Bank accounts are incredibly sticky, and even though it's not really that hard to open up a new account (and maybe switch over some recurring transactions), people nevertheless don't do it.


paying a lower rate literally solves the problem of having too much deposits.

they wanted a higher return so they increased their risk and blew up.


Well, yes, if your bank is receiving too many deposits and you didn't want to be vulnerable to a SVB-style failure, then you could:

- set your rate of paid interest on demand deposits quite low - what's going to happen, some people will pull their deposits? That's fine, that's what you want.

- re-deposit those excess deposits at other banks, taking only their meager interest payments on demand deposits

The math here works fine. As long as there's some difference between the rate you're paying on deposits and what you're getting, no matter how small, you're fine. And if you need cash quickly, hey, those are DEMAND deposits at other banks, you should be able to withdraw them immediately. You can spread the risk of a run on your bank around to every other bank.


The customers who deposited $150 million expect some rate of return on their deposits - in fact, you promised it to them. In a year that $150 million needs to turn into $155 million or whatever.

So you need to lend it out, and charge interest, and use that interest income on your loans to pay the interest on your deposits.

Sounds like SVB made a lot of loans or investment purchases quickly, and then some of those went bad.


Is this not what bonds are for? Savings accounts earn less than just buying bonds, and the bank pockets the delta.

They got greedy.


That's exactly what the bank did - they bought bonds. The type of bond they bought doesn't matter, what matters is that in a rising interest rate environment, the face value of the bonds fall and they are no longer producing higher yields than the saving account rates.

They didn't get greedy, they failed to anticipate the speed and amount that interest rates would rise.


Yeah, but they bought bonds to make their numbers work. They had a choice of getting like 0.08% in overnight funds or 0.36% in short term T-bills or locking it up and getting ~3% in MBS. They chose the latter to make more money. If they had chosen the former things would have been far less critical.

In retrospect this was a huge risk. They locked up way too much money in long term securities for how flighty their deposits could be.


3% MBS, that's a good one. Try more like 1.25% coupon rate with the crazy low mortgage rates were at when SVB got the inrush of deposits (the banks and the GSEs gotta make their money too, with mortgage rates near 7% UMBS coupons are only at 5.5%).


They too, thought that the world had fundamentally changed as a result of the pandemic.


The type of bond would matter if default risk was a meaningful piece of the puzzle here, but in this case it isn't (IIUC).


Bonds have a duration until maturity. Savings accounts can be withdrawn at any time. The bank normally pockets the delta by taking up the risk of this mismatch. In this case the risk became too much.

Think about it from the other end. You have a mortgage. But the bank needs money now and asks you to prepay your mortgage. What do you do?


I can't edit this anymore, but: there are multiple articles explaining that SVB owned mortgage-backed-securities, which is what I meant by "getting greedy".


They did not get greedy. They bought bonds, and those bonds turned into a liability. Because the bonds were bought when interest rates were extremely low, they are worth less than bonds at current rates and had to be sold at a loss in order to shore up liquidity. That spooked investors and prompted a run on the bank.


You said they didn't get greedy but the next sentence is the description of an extremely greedy act. They did not need to buy 10 year bonds. They could've bought 1 year bonds, or any other length shorter than 10 years. Or just less 10 year bonds.


> They did not get greedy. They bought bonds

they got some bonds which had an interest rate risk, in order to earn a higher interest. This is "greedy", but it would've been fine under normal circumstances, since they did not break any banking regulations.


The bank has short-term liabilities to their customers, and decided to lend out the customers funds to someone else for the long term.

Total non-sense and greediness.


I believe it's because depositors want a return on their "investment"


Because the customers expected interest and the bank’s investment portfolio can’t provide the necessary return to pay the interest rate they used to attract the customers.


with interest being zero as it was the past couple of years, can the bank not just sit on that money and literally do nothing? What operational expenses do they have?


Sure, if depositors said "ok we'll let the bank ride this out." But that's not what happened.


Yeah, and remember that this particular bank has a very, very well-connected set of customers (in the graph theory sense). So once the run starts, everyone tries to get in, causing the actual run.


rent and maintenance if they have branch property, salaries and benefits for staff at least. Banks have a lot of regulation and audit requirements which take work.


but that doesn't increase with more deposits, right?


if profits don't go up, shareholders get mad and the stock price tumbles. this is just capitalism.


Can you ELI5 this for me? Why is it terrible for the banks to simply sit on deposits if there aren't sound lending opportunities? I understand that the owners might prefer to sell rather than wait, but a small bank doesn't seem any inherently different than any other small business who's owner simply enjoys the work of running something. What is the inherent downside to idle deposits?


Deposits aren’t free. We will use round numbers. If I take $100B in deposits and pay .5% interest, that’s an expense of $500M per year and including overhead even when rates were 0, deposits at a bank cost more than 0.5%.

So I need to lend it out so I can make the spread and I’m not just bleeding it money


Deposits don't need to have interest paid. It's entirely possible for the money to simply sit there and not be invested in anything.


Four questions - where does a bank store cash? What is the IORB and what are its underlying assumptions? What is the overhead cost in banking? What is the cost of ACH and a wire? I don’t contend you can just have money sit there. I also don’t contend that it was stupid in hindsight to be this levered to duration. But there is an overhead cost in banking.


Yes, it's nonintuitive at first glance. A (good) loan is an asset, and a deposit is a liability as far as banks are concerned.


>Until I heard about those banks I hadn’t considered “too much money” was a problem.

You can always buy crypto if you think you have too much money. Or find a nice Ponzi scheme and invest in that.


> “too much money”

Its not so much this as deposits are bank liabilities.


> - The issue is that as the Fed raised interest rates in 2022 and continued to do so through 2023, the value of SVB’s MBS plummeted. This is because investors can now purchase long-duration "risk-free" bonds from the Fed at a 2.5x higher yield.

Let's be clear, the issue wasn't that the Fed raised rates to a historically average level, it was that they were manipulating the bond market in 2021 with trillions of dollars of QE.

Over the last few years the Fed has basically done a pump and dump on the bond market, and SVB being a bank was basically forced by regulation to buy long-dated bonds for yield.

I've seen a lot of people speak critically of SVB and I get it, but I think people should take a minute to ask why the hell bonds were yielding such a low amount in 2021. I just wonder how much longer we're going to blame, banks, crypto investors, bond investors, equity investors, home buyers, etc for what's happening to the value of their assets. When central bankers make government bonds trade like meme stocks this is what happens. Perhaps if we didn't do that, SVB and many others wouldn't be in this position.


>Let's be clear, the issue wasn't that the Fed raised rates to a historically average level, it was that they were manipulating the bond market in 2021 with trillions of dollars of QE.

One of the principle, statutory purposes of the Federal Reserve is to conduct monetary policy to achieve maximum employment and stable prices. That means it's the job of the Fed to manipulate interest rates.


Sure it is their job to set interest rates. It is far less clear that it is their job to buy government and corporate debt, and allow the government to issue unprecedented amounts if debt. Or to attempt to completely erase the business cycle.


That would be the Fed pursuing the "maintain the stability of financial markets and contain their systemic risks" part of its statutory purpose.


From current evidence, their efforts seem to have had the opposite effect.

These actions which the fed has never before undertaken (qe and zirp and even buying corp bonds) were to blame for the bubble, and the problems caused by unwinding it.


Covid would have forced a much harder reset, probably in 2021.


Covid came at the end of a decade of QE. They tried QT in 2019 before Covid hit and had to give up.

QE and ZIRP is the original sin here which created the dilemma the fed now faces - hard inflation or hard recession. They didn’t solve the GFC or Covid, they just postponed the impact and made it far worse.

I have absolutely no objection to counter-cyclical monetary policy but the monetary policy has become the cycle.


mind you the alternave is far worse.

Having unstable prices for staple goods will lead to unrest very, very quickly, which in term results in a downturn in the econonmy, which in term leads to even more unstable prices and thus more unrest.


Actually, they are sacrificing employment to stabilize prices: https://time.com/6253699/federal-reserve-inflation-interest-.... The interest rate hikes are designed to cause unemployment, and it's not even working.


Well, yeah, their "dual mandate" is contradictory. You're not the first to notice, trust me. And that means that sometimes they have to sacrifice one for the other. When one is doing historically well, and the other not so great, it's probably not a hard choice.


You’re inferring good faith, which - when dealing with humans - is an unsupported assumption. You don’t have to be an economist to draw the connection between printing money and a rise in nominal prices (irrespective of the interest rates and lending).

The Fed knows that there is no real ongoing inflation. The devaluation of the dollar already occurred and the new price has to propagate through the market. Their actions have no effect on the cause or broader course of apparent inflation only on who “wins and loses.”


>The Fed knows that there is no real ongoing inflation.

Can you cite any metric showing there is "no real ongoing inflation"? CPI, PPI, PCE, and other less-commonly used metrics all indicate ongoing inflation. It would be interesting to understand how you've arrived at the conclusion there isn't ongoing inflation.

>The devaluation of the dollar already occurred and the new price has to propagate through the market.

The DXY has been uptrending for nearly a decade, even more rapidly so since mid-2021.


They have a dual mandate. Pricing stability and employment. They're trying to trade one for another to achieve a better balance. Employment is far too hot right now.


It's a dual mandate because it's impossible to accomplish both. They have to decide which to focus on.


Don't see how anyone could view the feds actions in the last few years and conclude that they had this as their mandate


> maximum employment and stable prices

It's a dual mandate. If I'm not mistaken, we recently reached pre-pandemic employment levels. There is no way we would have reached this point without low interest rates through the pandemic.

Other countries have the same inflation rate that we do, but with lower employment rates. Ours is a better position to be in.


the economists at the fed think the only possible cause for inflation is demand pressure, forgetting the historical supply crunches that have led to inflation. wages are up and unemployment is down because production is still reeling from covid, especially as the largest generation in history retires. so amusingly, in their failed ploy to crush labor power, the feds have fucked the banks.

(this is not to excuse the greed of the bankers - this crisis is the purest essence of capitalism, it's inherent contradictions on full and gory display.)


> the economists at the fed think the only possible cause for inflation is demand pressure, forgetting the historical supply crunches that have led to inflation.

No, the economists at the Fed do not think that. Or rather, they try to find out what is driving inflation in any given situation, like they did with this 2022 study:

> Inflation has remained at levels well above the Federal Reserve’s inflation goal of 2% for over a year. Separating the underlying data from the personal consumption expenditures price index into supply- versus demand-driven categories reveals that supply factors explain about half of the run-up in current inflation levels. Demand factors are responsible for about one-third, with the remainder resulting from ambiguous factors. While supply disruptions are widely expected to ease this year, this outcome is highly uncertain.

* https://www.frbsf.org/economic-research/publications/economi...


It is more like, demand is the only thing the Fed can control.

The Fed can’t end the invasion of Ukraine or cure avian flu or speed up cargo ships. They can raise rates and that’s about it.


Nah, the bank is responsible for their decisions.

They bought $80b of fixed-rate bonds at historically and artificially low interest rates in a time of massive QE. Even based on the information available at the time, this is not a surprising outcome at all.


And the $80 bill was about 40% of their assets. And their depositors are all businesses who will move the money out fast because it's not insured over 250k.


Yeah, this is all very odd. What the hell happened there?

In retrospect, the Chief Risk Officer leaving and not being replaced immediately was a bad sign: https://fortune.com/2023/03/10/silicon-valley-bank-chief-ris...


That's a good find...

100% speculation: the CRO told them loading up on long term bonds was crazy, and that they'd just have to eat lower earnings. The CEO/CFO disagreed, and the CRO didn't want to have any part of it. Hindsight is 20/20, but... this really was a disaster waiting to happen.


Not completely sure about that:

"A risk officer typically anticipates and manages regulatory, operational, competitive or other risks faced by a firm."

That looks like _non-financial_ risks. If that was the role of the CRO at SVB, the risk for this event probably should have been with the CFO?


“Operational” at a bank includes financial risks.


I know of at least one large European bank where it explicitly doesn't.


that doesn't excuse the Fed's behaviors.


It's easy to criticize the fed, but they did a pretty good job during COVID and they did a pretty good job during 2008.

The charge of "manipulating bond markets" is pretty absurd - their mandate is to influence inflation and employment via interest rates. Of course it has an impact on bond markets.


Not to let the banks off the hook, but

>They did a pretty good job during COVID

One word: "transitory".


And it was. The prediction was that covid supply/demand constraints would resolve when covid problems clear themselves.

Unless, of course, you are expecting the Fed to predict the war in ukraine, the energy crisis, etc.

And in any case, people who take the Fed's predictions at face value, and base their financial positions on it, can only blame themselves when the Feds do something contrary (as they are always entitled to). These predictions aren't promises.


It was surprisingly accurate.... mom inflation has been reasonable for several months now


By that metric, and it took quite a bit longer to get there than ideal. It's also still elevated historically and compared to FFR, all of which is a roundabout way of saying that inflation has and continues to destabilize the outlook. Stability historically comes after an n-month lag of the FFR and core inflation meeting, right? Well, banks are failing now, and 4Q24 is a ways off.

I just don't know if I'd call it a "pretty good job," is all. They were caught-off guard and didn't move quickly enough.


They are always caught off guard. Try predicting anything macro and see how well it goes to plan. Then try not moving too fast or too slow.

The idea that they aren't doing a good job because they aren't perfectly precise with their timing is misguided. It's a crazy hard job.


The difficulty is somewhat artificial. Their statements give away the game, namely their obsession with a "soft landing." Translation: an undue focus on securities markets and saving entities who'd made bad investments, i.e., picking winners. They were late because they were trying to balance a factor that isn't part of their quasi-mandate (but that is important to politicians, and to the institutions that Fed officers revolving doored from). Without that factor, you see radical hikes in 2021 and QT not being undermined by waves hands vaguely. It hurts, but less than what's coming now.


The only entities they seem to worry about is banks. And banks going under is bad news for everyone, always has been. Being a lender of last resort to banks is explicitly part of their mandate.


They worry about banks because they are a bank. If it weren't for government oversight they'd be a private bank.


>The only entities they seem to worry about is banks.

You'll have to support thay with something. Again, the language of their correspondence over the past few years says otherwise.


That's.. Delusional. It's barely budged.


https://www.bls.gov/news.release/pdf/cpi.pdf

Note all the months since July 2022. It's good to read the source material rather than hyperventilating news articles.


Mean MoM % change in CPI-U from January to preceding July, 2016-2023:

* 2016: 0.01

* 2017: 0.20

* 2018: 0.27

* 2019: 0.10

* 2020: 0.21

* 2021: 0.30

* 2022: 0.61

* 2023: 0.27

Raw monthly data: https://data.bls.gov/timeseries/CUSR0000SA0&output_view=pct_...

I was skeptical that month-over-month CPI had come down to reasonable levels over a wider time window, but it is much lower than the same period a year ago, and close to recent historical norms outside that inflationary spike. I am eager to see Tuesday's numbers for February inflation.


> One word: "transitory".

It has been. We're barely at 1.5-2 years in length and it looks to be either stabilized and perhaps dropping.

A number of folks estimated that the COVID recovery stimulus would take about this long to work its way through the system, using Korean War spending as an analogue:

* https://www.piie.com/blogs/realtime-economic-issues-watch/in...

The fact that energy and food prices spiked due to geopolitical events threw a monkey wrench in the works temporarily, but now that those have stabilized, so has inflation.


The magnitude matters, too. Stimulus did not constitute an annualized increase in average consumer purchasing power equivalent to the inflation we've seen. I imagine that wage inflation fills the difference, but of course, the stimulus is over and wage growth is not enough. So "transitory" is not an accurate characterization for most people looking at their finances; it has come to eat their purchasing ability, and stayed.


To say they did a good job during covid can only be determined through history… many would disagree.

They acted as aggressively as possible to remedy a short term problem, which just created longer term problems. Its easy to throw money at a problem and worry about the consequences later.

SVB wouldn’t have failed if the Fed hadn’t artificially suppressed rates to the extent they did, for example. And Id venture there’s a lot more to come on the economic front. Many assets became extremely overvalued which makes the larger downside on price more disruptive… commercial real estate is a big one to watch in the coming years

If inflation remains entrenched we may require a GFC style hard landing to quench it.

The fed did head off an immediate disaster through suppressing the rapidly widened credit spreads, but everything after that will likely be remembered as a failure


They did a terrible job after the Euro sovereign debt crisis passed, so let's say after 2014, and they did a terrible job starting mid-2021, after vaccines and opening up. Always too slow to raise rates and stop financial froth that was obvious to everyone.

The first half of stopping a panic is always easy if you're the Fed with the exorbitant privilege to print money with virtually no immediate ill effects. The second half is the hard part and they've failed 11 times out of 10.


The fed wasn't buying bonds for lulz, they were doing it to carry out their primary function.


What's the fed's obligation to that bank that you are making this point?


Dollars to donuts they’re a “sound money” kook who wants the Federal Reserve not to be able to manipulate the money supply to achieve societal objectives like “low unemployment” and “inflation that closely tracks growth in economic activity and population” and “preventing a deflationary spiral because economic activity shut down due to a global pandemic” because all of that is interference with the Holy Free Market.

Because after all the purpose of society is to serve the market, not the other way around, right?


Two wrongs don't make a right!


Whether or not the Fed's behaviors are excused is not the topic of conversation.


... housing prices have continued to rise at wildly unsustainable rates, leading to record homelessness. Which is the exact opposite of what happened with the crypto market, where the Ponzi scheme collapsed.

When the assets haven't even moved in the same direction, I don't know how you are going to blame federal policy to counteracted the recessionary impact of covid for moving them. Whereas the fed has caused significant damage & also been ineffective at fighting the current supply-side inflation caused by Russia's war of aggression.


There's always degrees of speculation that are going to vary - cryptocurrencies, tech stocks, consumer staple stocks, gold, commodities, real estate, etc.

No one expects that real estate will skyrocket and tank like shitcoins (even in 2008, the residential real estate market didn't bottom out until 2012). But that doesn't mean that Federal Reserve policy has been good.

I'd highly recommend the book (or audiobook) "The Lords of Easy Money: How the Federal Reserve Broke the American Economy" for anyone who is interested in a historical summary of the FOMC.


Is there a reason they couldn't have just purchased shorter term bonds and securities instead?


Greed. Those were yielding close to zero and they wanted the 1.x%


It's an interesting reward mechanic for the bankers, "heads I walk home with big bonuses, tails I go home with a big severance and our customers get screwed".


This is the essence of the moral hazard created by the bank bailouts in 2008


The yield has been inverted so short-term has been paying just as much or more than the long-term depending upon which securities you're looking at. 30 to 90 day treasuries have been yielding over 4%


Yes-- if they kept that cash until now they would be sitting pretty


> SVB being a bank was basically forced by regulation to buy long-dated bonds for yield.

Is this true?


nope, there's no mandate to buy for yield - they do it to make money.


> When central bankers make government bonds trade like meme stocks this is what happens

That’s what happens when people buy meme stock. I think the failure of the goverment and the bank are unrelated here.


[flagged]


Or because it’s actually a huge scam.


“Crypto” is a technology that is functional and has proven utility. Those using it as a tool to skirt regulations, defraud others, or to commit outright theft are the scammers. I would agree that as a concept it is generally overhyped, usually by those who seek to profit from doing so. But I think it’s important to not conflate the unscrupulous gold miners looking to get rich quick with the utility of the precious metal itself.


> and has proven utility.

It hasn't.


The ability to have assets that can't be frozen by a bank or government beaurocratic fuckup is utility.


Not really because the assets can totally be frozen by the exchange. The people who have money in crypto that disappears don't care if that was due to a government action or due to whomever is in charge of enough servers.


bitcoin's main utility seems to be using an utterly absurd amount of joules per transaction.


And because crypto is garbage.


"This is not a liquidity issue as long as SVB maintains their deposits, since these securities will pay out more than they cost eventually."

But that's exactly the problem. With higher interest rates, those deposits will be looking for a higher deposit rate. With their assets tied up in low-paying long-term bonds, SVB will not be able to pay that higher rate.

It would only work out "eventually", if the depositors would accept a below-market rate until the bonds matured.

EDIT: This is indeed a solvency issue, not a liquidity issue, as also pointed out below.


He meant solvency issue. Someone else called this out downthread and he confirmed. Definitely felt like it should've been corrected more prominently, though.


And yet he is wrong!

They were stuck with assets earning <2% for 10+ years, meanwhile depositors were demanding higher returns (else they would put their money in other banks, in 6m tbills, 10yr treasuries, etc.. all of which paid a lot more).

And thus, "it's not a solvency issue as long as depositors are stupid and don't realize they can get more $$$ from other banks" ends up being quite misleading


Yes, I mis-spoke. If only there was some way to edit tweets or to pin your replies to the top of a thread :P


If you reply directly to your initial tweet with your correction (perhaps QTing the person who asked), I think it will appear at the top, or at least get higher ranking. I often see replies "boosted" via the original author QTing them as a reply this way.


So they didn't require the deposits earning 2% to be in term CDs like sane banks and credit unions do?


It's never a liquidity issue as long as no one tests the liquidity!


All liquidity issues are solvency issues when marked to market.


That's true in a tautological way, but there really is a difference between liquidity and solvency.

If a firm's assets will eventually mature and be worth more than the current liabilities, then a private rescuer can make a lot of money by bailing them out. If the assets will never recover or pay out, then someone will be holding the bag.


We've just seen there is no difference with SVB, because they tried to get a private rescuer to bail them out. Didn't work. Because why would a private rescuer lose liquidity to a lower paying assets (SVB's book) when they could lose liquidity to a higher paying one on the market (e.g. Treasury bonds)?

Mark to market is real. If you need liquidity immediately from the market, whether through a rescuer or otherwise, you do your accounting using that value and in SVB's case they were insolvent.


I think that is a difficult call to make this early on. The FDIC's job here is to recover value for depositors and find buyers for its assets. It is hard to organize bank takeovers on short notice, and even in a private rescue there is a lot of regulatory involvement. The story isn't even close to over yet.

There are some possibilities. Perhaps there was no single buyer that also had the (potentially hundreds of billions of dollars of) liquidity you would need to rescue the bank. Perhaps there were buyers who could come up with that liquidity reasonably quickly (on the order of a week or a month) but certainly not in the 48 hours it took for the bank to be completely vaporized. SVB was intending on raising liquidity from the market, and just the act of announcing their intent to do so triggered the bank run in the first place.


It think this stretches eventually. I think it as, If the current value of assets is less than current value of liabilities, we have solvency issue. If the current value is higher but will need some time to offload those assets, it is a liquidity issue.


Classically, it’s a liquidity issue if the market value of your bonds goes down because rates went up, causing the discounted present value of the future maturity payout to decrease. It’s a solvency issue if the market value of your bonds went down because your borrowers turned into smoldering craters and they will not ever pay you back 100 cents on the dollar.

Really it is not so obvious to discern between them at the moment that the crisis is at its worst.


Except SVBs assets would eventually mature for more than liabilities but no one would bail them out because the opportunity cost meant the book was worth less than just buying 10 year treasury bonds.


But it's important that they are eventually worth more. Enough to provide the rescuer with a substantial gain that compensates them for their troubles. Solvency in a "breakeven" sense is not enough.


“Eventually”

The markets can stay irrational longer than you can remain solvent.


One of the differences between a central bank and regular bank is that the regular banks should do the riskier stuff and offer the higher interest rates.

This in theory creates a diverse non-correlated system of capital deployment with the best projects winning over the bad ones.

However when the central bank offers interest rates that a private bank cannot match even when it's deploying into safe and endorsed assets like MBS then some weird stuff happens...

The Fed can promise risk-free returns at whatever rate they want but once it exceeds the private banks', then the banks no longer serve any purpose. If there were a way for individuals to hold accounts directly w/ the Fed, they'd all do that. Money will be sucked away from banks that deploy capital in the private sector and squeeze into ones that just passthrough to the Fed's like money market funds.

With high enough interest rates, the Fed can end up sucking up liquidity even from good and safe projects and cause widespread asset collapse b/c the entities that are supposed to be doing price discovery can't compete anymore.


"97% of these MBS were 10+ year duration, with a weighted average yield of 1.56%."

I'd like to learn more about the dramatic drop in MBS - elsewhere, downthread, it is asserted that they have dropped 30-50% ?

I understand the inverse relationship between bond price and yield ...

... but I am surprised that an asset yielding ~1.5% drops 30% in value when treasuries of similar duration rise to 3-4%.

Are there other factors at play with MBS in early 2023 - such as increased delinquencies - that are putting downward pressure on their price ?


I will first try to explain with simplified numbers and then do the equivalent calculation somewhere else.

Let's assume you buy a $100 bond with 0% rate for 10 years. For simplicity, let's assume it's a riskless bond. Now, let's suppose those same bonds now pay 5% a month later. Well, the smart thing to do would be to sell your 1-month old bond and buy the new one. Of course, everyone is doing the same thing so the price has to drop until the bonds are equivalent. You would get $100/1.05^10 = $61.39. Of course, the old bond still pays 0% but now, on paper, the price of the bond should grow 5% every year as we get closer to maturation.

Going to the real world, it would be something like 4.5% (current Fed rate - expected to be higher) minus 1.5% (the MBS rate they have) is 3% difference and so $100/1.03^10 = $74.41. Now, you said 10+ and so doing the same thing with 15 years is $64.19. This is also not including the fact that MBS is strictly worse than treasuries in terms of riskiness and so it's easy to imagine 50% off.


> You would get $100/1.05^10 = $61.39.

Emphasis on the exponential decay relationship between market price and time-to-maturity. If you change that 10 to a 30, the bond is worth $23, and if you change it to a 50, it's $8.72. For a bond that pays $100 at maturation.

I think laypeople intuitively guess that long-dated bonds are safer, because that is sort of how it feels when you are borrowing money to buy a house. But in terms of the market value of the bond, you add exponentially more interest rate sensitivity as the time-to-maturity increases.


The rule of thumb is that for each 1% increase in interest rates, a bond loses its years to maturity as a percentage.

So a bond with five years to maturity would lose 5%, and a bond with ten years to maturity would lose 10%.

If SVBs bonds had a ten year maturity when purchased, that's probably nine years remaining when interest rates increased from 1.5% to 5.0%, so that's 9 * 3.5 = 31.5% reduction in value.

If they bought $80 billion of these, that's a $25 billion dollar loss.

If their shareholders equity was $16 billion, then that's zeroed and a combination of depositors, the FDIC (for insured deposits) and other creditors would have to eat the remaining $9 billion.


It's pretty straightforward -- if you have asset yielding 1.5% forever, you can make it an asset yielding 3% by cutting its price in half.

In this case it's a little more complicated since you also get the principal back in pieces, but you can calculate the price today to create the equivalent of a 3% yield instead of 1.5%, and you'll get a significant price reduction.


There’s a misconception that bonds are safe investments. They are not. You’re just trading one kind of risk for another. You can do the math, compare 4% and 1.5% compounding for 10 years and that’s why no one wants the bonds yielding 1.5%. Dumping 90%+ of your liquid funds into a single thing other than cash is completely insane especially when it’s not yours.


Treasury bonds are "safe" in the sense that you will (because the US Government will not default on her debt) get your money back. The caveat is you will get your money back at maturity; if you need it before then, well, market value adjusts based on current yields.

If you're investing in bonds without building a ladder you're honestly doing it wrong. With the past 15 years of easy money and low yields it might have seemed pointless given rates barely moved, but completely giving up on any ability to capitalize on higher yields if rates move up is just poor investing :/


Bond ladders are the CPU pipelines of finance ;)


Bonds are perfectly safe investments when the normal consideration of safety is that you cannot lose money and you know your exact return through maturity. Can you miss out on better investments, ofcourse. The only issue is investing someone else's money into bonds - because they are the ones to decide when the cash is needed, not you. But I'd be shocked if at any given time at least 90% of cash is not invested in someway. You only want to keep what you need immediately out of investments.


Safe investment means you're not risking losing the money, not that there will never be a better opportunity (that may be just as safe). Alternative cost is not really coming into play here IMO.


MBS have a double whammy when rates rise. Not only are the mortgages yielding less relative to current rates but prepays decline so the duration extends. A lower relative rate for a longer time means a lower market price.


Interest rate rise will cause the same loss for bonds of the same duration. However MBS don’t have fixed durations. As interest rates go up, borrowers are expected to hold onto their mortgage longer. So the expected duration goes up for an MBS as well and loss is expected to be higher than bonds with similar duration to begin with. However there is a mitigating factor for MBS in a rising rate environment, that is they are amortizing instruments. So the duration doesn’t go up as dramatically as callable bonds/CDs.


Go watch Margin Call.


"to generate the yield they wanted to see on this capital."

Sticking to unrealistic goals seems to to be the downfall of a lot of financial institutions (and probably a lot of other companies). Same happened with Deutsche Bank in the 2000s. The CEO declared that they wanted to achieve higher ROI and to achieve this they had to start doing ever riskier stuff until it blew up in their faces (and the taxpayer generously bailed them out so they could keep their big bonuses).


Exactly that. I am wondering why the “pressure” to generate more yield or any yield at all. It is counterintuitive to me to view the startups deposits as investments and not as saved money to be used later, hence no need for a risky or any yield (even if part of the savings will be washed by inflation year over year)


Why would a bank take your money and go to the trouble of managing it if they didn’t plan on using it for something? The purpose of a bank is to lend money. They need deposits to do that.

In this case you’re sort of right in that they found have hedged a bit more by diversifying into more shorter term assets.

Hindsight and all.


I think there is a little more to it. I don’t know about SVB but Deutsche Bank had a solid business doing pretty unexciting banking. Then the CEO got a big ego and pushed for larger returns which made them do more risky stuff and eventually fail. So it’s basically a big ego or just plain greed


It's not clear under what conditions the startups gave money to SVB. It could be they just put into a checking account and SVB then invested it in a risky manner.


(We detached this subthread from https://news.ycombinator.com/item?id=35097120, which contains "that tweet".)


Great moderating dang! Appreciate your tireless work to make hackernews great.


Noob Questions: How do banks typically diversify their investments so that this kind of thing does not happen? Also don't they have to have some kind of liquidity cushion? Can't they just cover their short term costs by borrowing(I thought there is an overnight facility for lending between banks to borrow at low rates)


With only a look at the summary numbers above, it looks like they tied up 40% or so to 10+ years. I don't know what the right percentage should be, if that much is going to be tied up in hold-to-maturity, you would expect it on a rolling basis which reflects the long term liquidity of your deposits.

On its face, such a purchase would only be done assuming rates and markets will remain the same. I wish I could say that accusing a bank of making such a naive purchase means that interpretation is wrong, but these banks keep doing things like this since it's always worked out before. I'm sure it's much more complicated, but sometimes that's because it should have been a lot less complicated.


I don't understand why you would lock up that much money for 10 years at 1.5% interest.

Did they expect interest rates to stay at 0 for 10 years? That is illogical.


They wanted to chase high-yields in low yield environment. The only way to do so was to take on more risk, specifically interest rate risk. They ignored the fact that it was risky because it was a "safe bond" and got nailed by it.


If that was the market clearing price, then to the extent that rate expectations set the bond price (which of course is debatable) this would have been the consensus view at the time.


You can't really capture consensus with a single number.

Maybe the market expects a likely range of 1%-2%. Maybe the market expects a likely range of 0%-5%. Those could both have an expected value of 1.5%, with much different levels of risk for this use case.


Especially illogical, considering it's not their own money that they locked, but customers funds (that they owe to the customer in the short-term).


> - 97% of these MBS were 10+ year duration, with a weighted average yield of 1.56%.

This is a pretty insane bet. Why didn’t they ladder the maturities to have a lower average duration and less risk?


As a bank, parking the money into long maturity bonds, especially when it's not your money, and your customer can take the money back anytime, and the current rates are 0% (so can go upward only...).

Sounds like an insane investment decision.


[flagged]


What does ESG have to do with this? Seems like your politically motivated to see the world through one lens.


How is that relevant at all?


Not one mention on their values page of being responsible financial stewards. Only virtue signaling.


Bond traders are generally old white male republicans


My pessimistic view is that bonuses were paid out on invested cash not on cash just sitting there. So they had to buy something to get a fat bonus.


Shocking portfolio design isn’t regulated given how much of banking is regulated.


They were probably chasing higher yields in a low yield environment.


honestly disgusted by the blatant PR moves by YC and Founders Fund yesterday in leaking their “advice” to their founders to get out of SVB

Very blatant weaponization of FUD to drum up deposits for their investments in Brex, Ramp, and Mercury.


Or just some, as it turned out, valid business advice. That being said, I would never let my investors choose my banks (as in more than one bank) holding my company's cash. And I definetly wouldn't use some not-to-big-to-fail, not international bank to hold my multi-millions in VC money, which is the only yhing keeping my company a float.


Giving the advice is not what I’m referring to. I’m referring to intentionally leaking the advice to the press so they run the story about how Peter Thiel is warning everyone which accelerated the outflow.


Ah, I wasn't aware of that. I am not the least surprised so that Thiel is, however, involved in this.


As someone who was considering using one of those "banks" in the coming months, this whole ordeal makes me want to stick with Chase, Wells Fargo, etc. Stripe integrations be damned.


If you are outside of B2B, you do not need Stripe. You need a solid business bank, ideally multiple ones.

And if B2B is relevant, well, have an accoubt, or multiple, at a bank with Stripe integration to handle customer payments anf refunds. And keep everything else at different banks.


Would it be better to keep good advice private?


It’s probably not material to the overall picture, but just for accuracy, the wording of this is inaccurate - “As deposits grew, SVB could not grow their loan book fast enough to generate the yield they wanted to see on this capital”

Capital has a special meaning to banks, and deposits are absolutely not capital from their point of view of the bank (they are from the point of view of the depositor). To the bank, deposits are on the liabilities side of the balance sheet.

A bank’s capital is only equity put in by shareholders and retained profits from previous years. This is important, because how much a bank can lend is only determined by the amount of capital they have, not the amount of deposits (since deposits are on the wrong side of the balance sheet for lending from).


There are some big parts to this story we don't know.

Yes, they sold the treasuries and took a bath. But if that was their best option, it speaks very poorly to the other "assets" they held on their balance sheet.

We may find out in the coming days that they had a big position in Silvergate, which went bankrupt yesterday, and they had to mark their position to zero, creating the need for liquidity.


I heard elsewhere they were extending loans to startups with pre-IPO shares as collateral, so yeah, that was their best option apparently.


Something like 60% of their assets were these bonds in December. They very well may have seen the other 40% withdrawn already by startups that can't raise anymore leaving them with this pile of 10 year 1.5% bonds.


> An explainer post [1] connected to that Tweet is something I found extremely informative (assuming it's accurate):

[1] - https://twitter.com/jamiequint/status/1633956163565002752

That tweet is unattributed verbatim from https://www.livemint.com/news/world/explainer-silicon-valley...

[EDIT:] See the thread, it seems that the story may have stolen from the tweet! Pretty shocking for one of India's biggest business publications (and with 2 million Twitter followers).


He is correct, but he is blaming the FED raising interest rates. The responsible is not the FED, but the negligent management of SVB that purchased such products because of greediness.

When interest rates raise, previously issued bonds lose in value because there are more attractive ones available.

It's like if they missed the Chapter 1 lesson about investing into bonds.


If you inspect the values of SVB here: https://www.svb.com/about-us/living-our-values

You can see that managing money responsibly is not one of their values.


So because they have one webpage giving lip service to diversity and the environment, they automatically don’t value their core competency?

I’ve seen a few conservative tweets trying to tie DEI and ESG into SVB’s collapse in a weird attempt to tie it to a cultural war. It just doesn’t make any sense. DEI and ESG are side shows for just about any company and probably pay for themselves as marketing and employee retention for a Silicon Valley focused bank.

The values “We keep learning and improving”, “We take responsibility”, and “We speak and act with integrity” would all play into managing money responsibly.


>So because they have one webpage giving lip service to diversity and the environment, they automatically don’t value their core competency?

Not enough to put it on the page where they discuss things they value as an enterprise. Seems kinda important to leave out. And speaks to the failure of competency.

They don't value merit, they value meeting arbitrary quotas to look good, and now they met their end. Totally relevant.


My tweet was posted before that article.


What are you saying?

1) You wrote the Mint article.

2) Mint took the text from your tweet unattributed.

3) ???


Looks like #2


> Looks like #2

Wow, this appears to be the case, at least judging from the timestamp on the first reference I could find: https://twitter.com/livemint/status/1634091187442266112

That's massive journalistic malpractice!


Some people are downvoting my reply to jamiequint https://news.ycombinator.com/item?id=35100305 but if that comment becomes dead then you won't be able to see his reply to me or the surprising conclusion of the story.


When the fuck are retail banks going to be ring-fenced away from being able to trade in MBS. They're consistently cancerous to our banking systems.


I do not think this is a problem of mortgage-backed securities.

The problem is that SVB tied up their liquidity for 10 years at a yield far lower than they would get with more secure investments after the FED's rate hikes.

The specific assets they invested into are immaterial.


Aye, the mistake was the duration, not the instrument. I hope nobody would take 60% of their brokerage account and invest it in a 10 year bond either, ladders exist not just to manage liquidity but also to limit the duration you have to suffer low yields with.


Yep. Let's not forget this same thing pretty much happened last year in the UK - pension funds got margin called because they borrowed money to buy gilts (UK gov bonds). Low interest bonds, money tied up, messed them up when interest rates rose.


Come on, where else would they put your money ? When are retail going to understand handling and parking and securing and regulating money has a cost and interest rate have to be chased somewhere.

Narrow banking never works because at the first regulatory lapse (my employer got fined 200M because we used whatsapp, not even for committing crimes), BAM no money left for deposit liabilities.

And dont forget here that as long as people eventually pay their mortgage in the expected default risk, the money will eventually come back, at an opportunity cost.


This has nothing to do with MBS in particular, it is a fundamental aspect of the fixed income market.


This isn't an MBS problem. It's just a poorly performing asset problem.


This is a symptom of the problem of middle class single family home residential real estate being treated as an unreasonably-price-increasing bubble inflated investment and not a place for people to live in.

The irrational exuberance in price increases in this segment of the market over the past 4-5 years is not sustainable.

It is not logical, sane or normal for houses that were valued at $150k five years ago to now be valued at $400k in some suburbs and metro areas.


It's logical or sane if you think capital investment options in New areas is going to dry up and existing assets are the best opportunity for preserving or growing your money.

How much would you spend on the house if you're only alternative is to watch your capital disappear


> How much would you spend on the house if you're only alternative is to watch your capital disappear

If I'm worried about 10% or 20% disappearing to inflation, I'm certainly not paying 2x on an investment to prevent that.


What if it is 10% per year until it is all gone?


How many people think inflation is going to go to 10% and stay there for a decade or longer? Are they enough to drive up real estate this much?


Maybe not 10%, but I don't think we will see deflation anytime soon so cash will always lose value. Sometimes faster sometimes slower.

Capital needs somewhere to go as long as money keeps being printed. If there is more capital than profitable investment opportunities, it will flow into housing.

As a thought experiment to illustrate the point, imagine a world where there are no investment opportunities and the only place to store value is cash or housing. In this case, housing prices will continue to climb because money keeps being accumulated and there is no place to put it.

Of course this is a hyperbolic example, but I think our world is moving closer towards it as truly valuable Investments dry up. Certainly in the short term, and possibly in the longer term as well



Bailout? Sounds like an '08 déjà vu again.

It's crazy to realize how brazen Wall Street bankers and Hedge Fund managers are when it comes to asking for taxpayer dollars during distress, while decrying any attempt to add to that taxpayer dollar pool from their billions in earnings when the times are good.


> "- In 2021 SVB saw a mass influx in deposits, which jumped from $61.76bn at the end of 2019 to $189.20bn at the end of 2021.

Would be interesting to see where that money came from. That has all the markings of a pump before a dump, the dump being 2023. They didn't even have to dump, they just had to stop the pump to auto dump once the interest rates went up.

SVB opened themselves up to an attack vector and one thing the banking industry likes to do now and again is consolidate and shakeout. That amount of inflow in good times can make you do funny things. The better way to go about it is be scrappy always, and expect the attacks.

There were way too many companies in this bank, it had too much concentration of startup/VC/PE money. Regulations will probably have to be made around this now more robust.

HBS is even realizing too much optimization/efficiency is a bad thing. The slack/margin is squeezing out an ability to change vectors quickly. This is happening from supply chain to credit to food and more.

The High Price of Efficiency, Our Obsession with Efficiency Is Destroying Our Resilience [1]

> Superefficient businesses create the potential for social disorder.

> A superefficient dominant model elevates the risk of catastrophic failure.

> *If a system is highly efficient, odds are that efficient players will game it.*

[1] https://hbr.org/2019/01/the-high-price-of-efficiency


Good grief. Simplifying this to “they took on tons of deposits and felt the need to make risky unchartered investments to retain their existing level of returns” is just so stupidly greedy. JFC, I think here of a ship with a heavy keel; only for the captain to order cargo many multiples the weight of the keel to sit above deck.

Why can’t so many folks who are doing well enough just act with basic common sense and integrity these days?


I don't get it. I'm no expert in finance but even I knew the fed wasn't going to stop raising interest rates because I had the common sense to know the fed would fail to trigger a recession by doing so.


When I set my ‘Hindsight Goggles’ to 100, I too saw that the Fed would keep raising interest after the initial rounds, because unemployment would stay low despite massive layoffs, somehow, and that prices would keep rising. And I am an expert in finance.


There still a war on, inflation wouldn'tve gone down until that ended. And you have to take the temperature of things in your daily life, its like Keynes said, people are driven by the "animal spirit", the market doesn't always (or ever) make sense, you have to get a feel for it and use specific data to support that feeling, not the other way around. Jesus am I the only one taking out more lines of credit with the expectation of an extended period of inflation? If everyone else is doing that (which they should be), the money supply will continue ballooning, the economy will stay strong and unemployment low.


If people keep borrowing everything will be alright? Society, the economy, doesn’t work that way.

The economic good times are when confidence, trust, risk taking are all high. That music stops when people lose confidence, the Great Financial Crisis, just like all recessions are fundamentally a loss of confidence. This causes people with funds to claw them back, and retreat into their castle, and wait for the bad times to blow over. This is what happens when the economy collapses. When prices and asset liquidity crash. When rates go up. When you can’t borrow anymore and when you have to pay back the money you are borrowing, because it’s in the fine print that they can ask for it back whenever they like. Then inflation goes up. Borrowing rates go up. Asset prices go up, unless no-one has confidence in that particular asset (eg stocks or property where those companies are based).

Consumer leverage is not a bulwark against recession.


Based on what you’re saying here, is it a good time to buy a first home? The interest rate and prices have been scaring me but maybe I’m looking at it wrong.


I, too, am a genius in hindsight. It's great to be in such rare company.


Surely in retrospect, it must have been obvious you would find such people despite their rarity?


You got me there.


I thought layoffs were way below trend? What layoffs?



Stripe stopped payouts to SVB bank accounts. Seems reasonable.


I hope it's accurate. What I should have said though is "This is not a *solvency* issue as long as SVB maintains their deposits", it's obviously a liquidity issue.


Nitpick: The Fed isn't selling bonds. Maybe they meant the Treasury?


Yea that bugged me too ha ha. The Fed doesn’t issue any bonds.


> - This is not a liquidity issue as long as SVB maintains their deposits, since these securities will pay out more than they cost eventually.

This is exactly a definition of liquidity issue... Owning enough assets but not being able to cover short term liabilities.

In opposition to not owning enough assets to cover liabilities.


Relevant Bits About Money article on FDIC and bank failure:

https://www.bitsaboutmoney.com/archive/deposit-insurance/

I should add: apart from 2008, bank failure has historically been quite rare


Even if SVB maintains their deposits, that IS a liquidity issue, it just isn't a solvency issue.


> > 97% of these MBS were 10+ year duration, with a weighted average yield of 1.56%.

But why didn't they just hold money at the Fed given that they are a bank and they can?

It's literally splitting hairs between what the Fed Fund Rate is and what they got on their MBS.

Explainer post says end of 2021 they made that trade, in March the Fed raised the Fed Fund Rate to 0.20%, and by April it was 0.77%.

Had they waited just 5 months they'd have got a better Risk adjusted return by just keeping the money at the Fed


If they could have forecasted the future at that point, they could have made even more money than that!


If you could forecast even 6 hours into the future, you'd be the richest man alive in weeks. Nevermind months.


Because people can’t see the future and you’re talking do we forgo $1b a year or not in that spread. It wasn’t the MBS that got them into trouble, they were supposed to hold those to maturity. It was something else they did that led to that forced liquidation


Yeah, but banks don't profit by holding money for months.


Please correct me if i'm wrong since i'm not a finance guy,

apart from the loss of $1.8bn and the delta of the interest - growth from their assets.

Isn't everyone still going to get their money back - the percentage of the loss that SVB has which should be less than 5-9%? Sure FDIC has to liquidate all the money from the assets and it takes time. But at least the impact is not going to be as hard as losing all the money like FTX or Maedoff in 2008 right.


I mean. If you’re a startup that just had $10 million locked out and can’t make payroll. It doesn’t matter that you’ll “get your money back”. You’re toast.


> But at least the impact is not going to be as hard as losing all the money like FTX or Maedoff in 2008 right.

The recovery rate for the Madoff Ponzi is 88.35%: https://www.justice.gov/opa/pr/justice-department-announces-...


Not if they’re forced to sell at a low. I don’t know how big are the losses though.


all other details aside - if you purchase $80bn of “mortage-backed” anything you deserve everything that will eventually come your way


In general people pay their mortgages. Just because a financial crisis has happened around it once doesn’t mean the asset class is inherently bad. It’s more regulated than ever, the problem was rising interest rates and not being able to keep their head above water with the position.


“in general” is not good enough when you are pouring in $80bn mate.

would you mortgage (no pun intended) your family’s entire financial health on something that works “in general”?


What’s interesting is that this bank failure as well as the failure of a neo bank Xinja in Australia despite taking in deposits both illustrate that it is not the case that banks make money by “lending out deposits” but rather by issuing loans and then attracting enough deposits to make their lending operations profitable.


> 10+ year duration, with a weighted average yield of 1.56%.

> the value of SVB’s MBS plummeted.

How much 'plummeting' did they do in numerical terms? Something with those kinds of yields doesn't sound like it ought to be a super risky asset. The mortgage lending market tightened up a lot after the great recession...right?


They're very safe assets, they just have a long duration which makes them really risky if you could need them to cover deposits.

To make things more straightforward, let's just compress it to a 1-year time frame vs a 30-day bond. So a $100 MBS at 1.5% would pay you $101.50 if you held it for 1 year. If you have $100 in deposits and a $100 MBS bond, you're "solvent". But what happens if after 30 days, your depositor asks for his $100 back? You either need to sell your MBS or find other money to pay them.

If you try to sell the MBS to pay that $100 deposit liability and interest rates are about the same as they were when you bought it, you'd likely get around $100 and things are okay. If however, rates have spiked since then (like they have here), investors can either buy your bond that pays 1.5% or they can buy a new issuance 1-year bond paying 4% or 5%, or perhaps a 30-day bond paying 1.5%. So you need to give them a discount in order to sell your bond -- in this case it might be upwards of 30% or 40%.

So if you sell your MBS, you'll only get $60 or $70 for it -- leaving you a huge shortfall that you need to makeup from your other reserves. If you could convince your depositor to leave his money in the bank until that bond matures, you'd be completely fine -- but the timing mismatch and interest rate spike just kills the bank.


I get how it'd put them in bankruptcy or whatever the precise term is for a bank. I'm just curious what it means in terms of people getting their money back. If their assets lost 10% of their value, I could see that being enough, combined with the bank run, to put them under. But if everything else gets sold off at 90 cents to the dollar, that's not awesome but it's not like "poof it's gone entirely" either.


> But if everything else gets sold off at 90 cents to the dollar, that's not awesome but it's not like "poof it's gone entirely" either

You answered your own question. People will very likely get most, or even all, of their money back, just after the gov is able to offload some of the assets. Problem is, if you're a startup, you can't just wait a few months for the cash to make payroll.


A bond with a 10-year yield of 1.56% has a price of $0.85 on the dollar. A bond with a 10-year yield of 4% has a price of $0.676 on the dollar. So if yields increase from 1.56% to 4%, the bond price falls by 21%.


That has to be an inflation adjusted yield, right? Why would anyone do anything remotely risky for such terrible returns? You can almost find government bonds with similar average yields.


Nope! I got a 2.0% mortgage in 2021 (no points or anything) and the bank then turned around and sold it to Freddie who paid them 1.7% (so the bank made a nice 0.3% just for originating the loan).

Then Freddie packed my loan and sold it to others for something likely to be below 1.7%...


>You can almost find government bonds with similar average yields.

Not a few years ago. T-Bills were paying like 0.1% in 2021[1]

[1] https://home.treasury.gov/resource-center/data-chart-center/...


And 10-year bonds were well below 1%.

https://fred.stlouisfed.org/series/DGS10


Weren't the rates on government bonds negative if you were a large investor like a bank?


You can't use your intuition about stock prices for bonds/fixed income. In FI, its all a numbers game. As rates go up, prices go down.


The risky part is (or more precisely has happened to be) the 10+ year duration, more than the ~1% yield over treasuries (which may be too low to compensate for the additional risk but is not what has brought the bank down).


I had the vague impression that after the 2007 crisis, banks holding retail deposit accounts were not allowed to invest in stuff like MBS, only investment banks (without retail accounts) were.


> I had the vague impression that after the 2007 crisis, banks holding retail deposit accounts were not allowed to invest in stuff like MBS, only investment banks (without retail accounts) were.

You are confusing the Great Recession with the Great Depression.

The 2007 crisis and subsequent Great Recession was contributed to by the 1999 repeal of that rule, adopted in response to the Great Depression; there were several efforts to restore it after the Great Recession, but none succeeded.


The Glass–Steagall rules?

I think the financial crisis ended in 2009, when at the time there was fear that Citibank would be the next to collapse. A rule change was made to allow banks to value "hold to maturity" assets at the purchase price instead of the current market value, so that they could avoid an insolvent balance sheet. The problem for Silicon Valley Bank was apparently that they needed to sell some of the assets and take the loss.

* https://en.wikipedia.org/wiki/Glass%E2%80%93Steagall_legisla...


fifteen years ago we had simpler explanations of why things are going to pieces. Nowadays everything is more complicated - but the results are the same...

(i think HN needs a black bar, we are all screwed)


Pretty much the only outcome when reprehensible greed meets company structures indistinguishable from scams and charlatans masquerading as leadership.


Lincoln Savings and Loan, LTCM, Bear Stearns, and now: SVB.


This logic is counterintuitive to me “As deposits grew, SVB could not grow their loan book fast enough to generate the yield they wanted to see on this capital.”. Startups are not depositing the money in SVB to invest it, they are storing it for future use. Why the pressure to generate yield and grow the loan book “fast enough”?

https://twitter.com/AhmadBaracat/status/1634293096639787008?...


There was no explicit need or requirement for SVB to buy MBS at 1% yield, yes. Poor management handling too much money.

They also could have hedged the interest rate risk. Likely there will be policy change as a result of this. Banks over some AUM requiring stricter regulations

The Fed is complicit in encouraging moral hazard through distortion of the bond market. Pretty much every crisis in the modern era is precipitated by fed policy from years earlier


"The yield they wanted to see on this capital" I imagine is some combination of money needed to run operations of the bank, interest paid on the deposits and profit.

They could have just stored the money in the proverbial vault. But if they do that, then they have to charge the depositors a fee to be a customer. And competition has pushed in the other direction.

And probably more importantly that whole "profit" goal.


Because if you lose 7% of capital on $180 billion in real terms that’s an awkward conversation with your boss?


Bank accounts were paying .1% interest at this time. SVB could've bought 1 year bonds and made .5-1% easily. That's a billion dollars...


They didn't think they were gonna sell the assets before maturity. In that case, buying those makes sense.


Because the bank, like all capitalist entities, exists to make a profit. Why else?


Can they make profit by charging customers a monthly fee in return for their services instead of investing the money?


The team making these poor choices at SVB should be criminally charged ... The tax payer shouldn't have to bail out banks.


> should be criminally charged

For what crime?

If someone unintentionally but ineptly writes code that crashes Amazon on Black Friday, did they commit a crime?


It's possible to be criminally negligent. If you took unnecessary risks then it can carry a criminal element.


MBS are literally one of the safest looking bets imaginable, most likely Silvergate blasted a huge hole in their books and set off a cascade of trouble


Long term fixed income at near 0 percent interest when you might have to give the money you used to buy it back is very risky, as we see here.


What's a safer investment?


Another bank will acquire the company and make depositors whole. Government won't actually do the bailout, just facilitate/force it.


So where does the lost money come from? Another bank just gives it up?


SVB's shareholders take the loss in the value of their shares (which were down >70% before the regulators stepped in).

What will probably happen:

- Regulators will sell SVB to a bigger bank at a greatly reduced price which makes it worthwhile for the buyer.

- Depositors will be able to withdraw up to $250k cash starting Monday (the FDIC insured amount).

- Funds above that will eventually be recoverable in phases with the first chunk available by the end of next week (because SVB does have billions in cash, the issue is with the long-term value of their investments not meeting regulatory requirements.)

- If the system works properly, it's highly likely depositors will get 100% back. It may just take a little longer.

- There will be an investigation of the bank and its board's Risk Committee. Things aren't looking great for the CEO and the former Chief Risk Officer who unloaded some of their stock but that's more of an 'insider trading' beef against those individuals (ie unrelated to the bank's failure). So far, in terms of risk management, it doesn't appear SVB did anything out of bounds or even unprecedented.

Arguably, as conditions changed last year they chose a particular investment strategy which would normally have worked out but instead hit a pretty unlikely perfect storm of external factors including the Fed raising rates really fast. We don't have all the info on who knew what and when but my guess is: if there's "wrongdoing" here, it's probably more stuff like not taking the corrective steps they took this week sooner. They may have been hoping to "play through" the rough patch and it could have worked - but this time it didn't because there was a run by their unusually close-connected depositors in the tech startup community.

(Note: some folks are arguing that the Fed itself created a volatile situation for this bank (and others) by doing pretty unprecedented things which rapidly impacted the bond markets and other securities tied to Fed rates. I find this argument not entirely unreasonable. We won't know root causes until regulators unwind this.)


Another bank will commit to meeting 100% of investor deposits, take ownership of all SVB's assets, and provide liquidity to shore up any depositor concerns. Why would they do this? They get to acquire SVB insanely cheap (basically just the cost of covering the losses), get loads of new now-happy customers, and be hailed as a hero.


Unless, of course, this predicted white knight never appears because lots of other banks are secretly in similarly shaky positions where they are also holding lots of long-term Treasuries or MBS that are in some respects extremely safe but which also pay only 0-1% over 5-10 years, and those “assets” would need to be fire-sold at 65% of face value if the bank ever needed cash quickly…

And even if that white knight or even off-white-ecru knight did come along and want to buy the bank next week, what if they don’t decide to make every depositor whole, for those holding cash above FDIC limits? They certainly don’t have to do that. They could just buy the loan book. Or the warrants or early debt for a number of Silicon Valley startups. They can be vultures, not Santa Claus…

I think people don’t quite grasp what could be coming next.


>And even if that white knight or even off-white-ecru knight did come along and want to buy the bank next week, what if they don’t decide to make every depositor whole, for those holding cash above FDIC limits?

The wisdom on the street is that the FDIC will not let that happen as it would potentially cause a run on every bank outside the top 5. If there is a bank large enough to make every depositor whole without causing stress, then the FDIC (on a 10 year time frame) is handing you $100B "for free".


> I think people don’t quite grasp what could be coming next.

Agreed on that. The FDIC will make the best deal possible for depositors, but they'll spend a very busy weekend figuring out what that maps to in reality.

We don't know yet.


> Unless, of course, this predicted white knight never appears because lots of other banks are secretly in similarly shaky positions where they are also holding lots of long-term Treasuries or MBS that are in some respects extremely safe but which also pay only 0-1% over 5-10 years, and those “assets” would need to be fire-sold at 65% of face value if the bank ever needed cash quickly…

Just by coincidence Wells Fargo is having "technical glitches" affecting account balances the exact same day this happens.


Wouldn't the FDIC take an enormous loss if it facilitated selling SVB "insanely cheap"?


FDIC is funded by premiums from participating banks. Beyond the insured limit, there isn't really a guarantee that the "lost money will come back."


They bought assets and loaned to customers … pretty standard stuff


> As a result, they purchased a large amount (over $80bn!) in mortgage backed securities (MBS)

Do we now have people making decisions on stuff like this who are too young or clueless to remember what happened with the 2004-2007 mortgage backed security bubble that popped in the 2008-2009 financial crisis? Seriously? Did nobody learn the lessons on this? Countrywide and other originators of MBS and CDOs?


Any investments whose value was sensitive to the Fed's rates would have had the exact same problem.

In 2008 MBSes were bad because the underlying value of the investment turned out to be bad. That's not happening this time. All that's happening is the same thing that happens to any bonds -- when rates increase, older lower-rate bonds lose value because why would anyone pay full price for them when they can get a new one with a higher rate?


> Did nobody learn the lessons on this?

I'm not sure you learned the lessons. Other than MBS being coincidentally involved, this has literally nothing to do with 2008.


Do you think MBS are always and forever a bad investment because of a bubble 15 years ago?

The MBS wasn't even the problem here. If they had 10Y corporate bonds or Treasury notes paying the same rate, they would have had the same problem.


> Do you think MBS are always and forever a bad investment because of a bubble 15 years ago?

I think heavy exposure to a single type of long-term asset for a retail bank is always going to be a bad investment decisions, because risks materialize, and correlated risks tend to materialize together, and when a retail bank suddenly lacks liquidity...


To be clear, I think MBS are morally wrong in how they're implemented in the market, yes.

A bubble 15 years ago? We're in a massive housing price increase bubble now.

Buying $80bn of MBS in the middle of a well-known housing price bubble is catastrophically stupid.


The underlying problem in the 08 collapse was poor/non-existent underwriting of the mortgages and those loans being rated AAA when packaged in an MBS. When the economy slowed just a bit people started defaulting because originators were writing NINJ (No Income, No Job) loans, something that is illegal today.

This situation today has nothing to do with the failure of the MBS'S to payout like 08.


The error wasn’t that the mortgages defaulted too much (like in the ‘08 crisis) but that interest rates went up, which is a distinct problem, and, from the comments in these discussions, not something that the capital requirements adequately capture.


The problem is with MBS but not for the same reason.


Hedging maturity and interest rate risk is literally the only thing a bank needs to do. (Okay not literally but come on you have more or less one job: assure assets are secure and generate some yield safely).


> - This is not a liquidity issue as long as SVB maintains their deposits, since these securities will pay out more than they cost eventually.

You mean it's not a solvency issue? It sounds like a textbook liquidity issue.


Why even chase 1.5%.


If you could make 1.5% off other people's money how much would you "invest"?


They got too much in deposits very quickly and could not originate loans at the same speed. If they kept the money uninvested their operating costs would have eaten up their principal (even if they had to pay 0% in interest to their customers)


but operating costs don't rise linearly with deposits (i imagine it rises sub-linearly, at most), esp. if they can choose to pay 0% interest for deposits.

It really is just a matter of money making. They didn't want to accept the low yield, but certain safety, and accepted a higher yield, but with risk. It caught up to them.


1% of 100 billion dollars is a billion dollars. They would rather have 1.5 billion than .5 billion.


On $80B that's $120 million a year


1.5% of $80B is actually $1.2B


Why would someone buy an MBS after 2008?


MBS is not the problem here. It’s not 2008.

They bought bonds when interest was low, now interest is high and they’re worth less.

SVB wouldn’t have lost cash on those bonds if they didn’t experience a bank run.


It is pretty predictable that they would experience a bank run given interest rates going up though. They locked the money up for way too long so it would've impacted the interest rates they could offer, meaning people would want to move their money outside of any panic instinct. Maybe if it were a bit more spread out over time they wouldn't have entirely collapsed, but they put 40% of their deposits in these things, it still definitely would've stung.

Not disagreeing that MBS is irrelevant, and yeah the risk rating of the bond wasn't a problem as this will pay what it said is would. Just saying it was a stupid move to buy such a large amount of such a long term bond when interest rates can't go anywhere but up and you're using borrowed money to do it. From that perspective it was a risky move.


It was stupid to not hedge that interest rate risk. I mean, the Fed telegraphed their moves well in advance. I don't work in finance but it doesn't take a genius to see how that dynamic would play out if that's your job. Not something I would've thought about till now but if banking is your business, you tend to think of ways it can break and this seems as elementary as anything if your job is to ensure solvency in all conditions. I just hope other banks weren't this stupid.


All banks still buy and sell MBS.


That sounds like a bad idea,but I am no investment banker.


The problem in 2008 wasn't really MBS it was the way they were packaged.

You take a collection of n mortgages and rank them by how likely they are to default, and then you divide them up into say 5 different securities each containing n/5 mortgages. The top 20% will be considered the least risky and the bottom 20% the most risky, and yields will reflect that. There are also rules on what risk level different asset managers are allowed to take on, so random retirement accounts wouldn't take the bottom 20%. No problem yet.

But then some "spherical cow" style math was applied to these things. If mortgage defaults are independent events then mathematically you can smooth out the risk even from the high risk ones by just grouping a bigger number together. So they took say 5 different shit tier mortgage collections, lumped those mortgages together to create another pool of size n, and repeated the process. Now the top 20% of that pool was given a very low risk rating, yet it still paid a greater yield than the original pool's top 20%, so why wouldn't an investor want to get in on this high yield safely rated security?

And it didn't stop from there, the bottom say 20% of those pools were again pooled together to create more supposedly safer securities.

This could've held up for a long time in a vacuum, as long as mortgage default rates didn't change too much. But part of what incentivized this system in the first place (and in turn what this system incentivized) was to hand out mortgages like candy basically.

So it was a house of cards waiting to fall, and once some of these securities faltered it also caused a bank run on a wider set of MBS structures that probably could've held out if not for the panic. Plus big failures like Lehman Brothers and Bear Stearns had impacts on firms that went beyond MBS activity, so there were some crazy cascade effects too.

Anyway, there's nothing inherently wrong with MBS, it's just important to know what you're buying. This situation is weird because it's not that the security itself is unsafe, it's the context in which they are using it that is risky. Though I suppose it relates on the high level that they probably thought "this security is rated very safe and it gives relatively high yield" without an actual understanding of the risk involved in their use case.


MBS is a critical part of the american economy. You wouldn’t be able to buy a house without it.


[flagged]


Because, as described in the link, the gap between deposits and liabilities is likely to be over 25B.


Because most of the potential investors are sophisticated enough to realize they wouldn't be 'investing' as plugging holes in what would be rapidly turning into a Ponzi scheme.


If that's even mostly true, it sounds like the SVB was set up to fail. It almost sounds like like Biden is trying to crash the US economy.


Investing in the exact same kind of unsafe assets that brought the 2007 crisis, as well as assets that cause house prices to stay unaffordably high.

Yep, all of SV is truly made of bumbling idiots. That whole solution is truly hilarious and watching all these clowns lose their money is going to be fun.


Actually not. The kind of asset they invested in does not really matter. If they invested in super safe government bonds at <1% at the same time, they would have had the exact same issue. It is the rapidly rising interest rates that did them in. If they were smarter they could have done a rolling ladder of short maturities but probably someone there was lazy.


They could have stopped offering the very high savings yields that attracted that capital, but profits and exec comp would have dipped. So...this instead.


Gotta love the monday morning quarterbacking. "Get rid of the thing that attracts all your customers, surely then there will never be a bank run."


I suppose you’re right, there’s nothing they could have done differently. They were forced to offer just about the highest, unsustainable, yield rates and otherwise attract and hold unmanageable amounts of capital. No other choice! Inflation was unforeseeable. Despite 40 years of startup banking experience and $3-10M yearly exec comps, how could they be expected to consider high inflation and a tech downturn happening at the same time? We are asking just too much of those poor execs, risk management must have been someone else’s responsibility. Biden must be to blame.


That would have resulted in just as much of a flight, to higher yield instead of safety.


No it wouldn't have. The company would still be solvent and have a large number of total deposits, just less in total than they could've had otherwise. What they did literally destroyed the entire company; not doing so would not have had that result. The outcomes are not comparable.


What? Anybody with financial sense has noticed that rates have skyrocketed, and if you're still earning less than 1%, you're throwing money away. The more money you have, the bigger a deal that is.


Bank accounts are very sticky and most people don't move them in search of better rates. In particular, people are not thinking of cash sitting in a checking account as anything other than short term cash they don't expect to earn a return on anyway, so you wouldn't change your entire commercial bank (that everyone else in your industry also uses and expects) just to chase some small returns.


SVB was holding so many uninsured deposits because they were the operating cash for businesses. If you have $100M, the difference in interest would be $3M a year between 1% and 4%. What kind of business wouldn't be paying attention to that?


Banks were not offering any interest on deposits at this time. Maybe you could find .2% if you looked hard


The execs have been pulling $3-10M/yr compensation and will walk away wealthy, legally in the clear, and probably into similar roles and comp elsewhere (or simply retired). They aren't bumbling idiots, though if they were, they'd be hella crafty ones.

I don't think the depositors, investors, or VC's that pushed for SVB were idiots. SVB was a good bank, with a good reputation, and good services that got mismanaged into oblivion.


I suspect all depositors will be made whole. The bank had a liquidity crisis; it had reserves in excess of its liabilities.

Every bank borrows short term (you can walk up and withdraw your money at any time) but lends long (e.g. mortgages, though SVB writes few of those). The recent management grabbed some very long federal bonds; as rates have risen the resale value of those long term assets (paying a lower interest rate) fell. They can't unwind that position and cover all possible demands.

FDIC will transfer the accounts to another bank, guaranteeing the 250K at least (I believe SVBs own liquid assets could cover that) and may use its own asset base to cover the balance, siezing SVB's other assets and stuffing them into FDIC's piggy bank. It's not like those government bonds won't pay out...eventually.


SVB held $21bn of 'available for sale' bonds and $91bn of 'held to maturity' bonds on its balance sheet, that were actually only worth $19bn an $76bn respectively on a mark-to-market basis, which means a total unrecognised hole its in balance sheet of $17bn. SVB's total equity was only $16bn[1][2]

That means it didn't have a liquidity crisis, and it didn't have reserves in excess of it's liabilities, it had a solvency crisis, and it has more liabilities than it has assets (before even considering the haircut it's assets will suffer at liquidation prices).

The crux of the issue here is that, for many types of assets, banks are able to test whether they meet capital requirements based on the price they paid for the assets, rather than the price the assets are currently worth. So SVB was sailing along nicely whilst it's bond portfolio slipped further and further under water, and wasn't required to recapitalise when it should have done.

[1] https://ir.svb.com/financials/annual-reports-and-proxies/def... [2] https://www.wsj.com/articles/bond-losses-push-silicon-valley...


People keep talking about how 'this is a solvency crisis because if SVB had to sell everything today, they wouldn't cover liabilities' when that is the definition of a liquidity crisis.

EDIT: To be clear, think of it this way. I have a piece of paper saying you'll give me $100 in 1 year plus 1% interest that I bought for $98.

No-one buys that piece of paper for $98 today, because they can get the same deal with better interest. But that doesn't change the fact that I will get $100 for it.

If deposits hadn't shrunk, $100 would go to SVB in 1 year and everything would be fine, it's the fact that they have to sell it so far ahead of maturity that's the problem, we just didn't notice this phenomenon in the past few decades b/c rates fell and prices went up.

This is not because SVB has a particularly risky book (we're talking treasuries here), it's because they didn't account for declining deposits (itself a very stupid, but unique bad decision unrelated to their risk tolerance).


> People keep talking about how 'this is a solvency crisis because if SVB had to sell everything today, they wouldn't cover liabilities'

This is not accurate, and the inaccuracy is the difference between solvency and liquidity.

If svb had longer (ie weeks), they still wouldn't be able to cover their debts. Its not a matter of needing time to arrange buyers for their assets; their inability to pay isn't related to liquidity today or tomorrow, it's related to their asset's value. If they snapped their fingers and marked to market all their assets, they would be in debt because they're insolvent.


> to pay isn't related to liquidity today or tomorrow, it's related to their asset's value

How do you calculate that precisely? That seems like a fuzzy distinction, considering that the value of things change over time.


> seems like a fuzzy distinction

Not really. Insolvency is by definition time independent. If you owe more than you have, you're insolvent. It doesn't really matter if you might make enough to cover the difference in the future.

In practice, you might get away with it if no one forces the issue upon you, but that doesn't change the math of whether or not you could pay your debts.

Liquidity is different. It is time dependent by definition. Some things take time to structure, deal, and sell. You can also get away with this in practice.

They can both have similar effects when they happen, but their causes are sharply different. I have a house I can sell to cover my mortgage, but couldn't sell it in less than a couple weeks or maybe even months. Illiquid but solvent.


Insolvency is a state of financial distress in which a business or person is unable to pay their bills.

If you think it is anything else you are just being an apologist.


> Insolvency is a state of financial distress in which a business or person is unable to pay their bills.

This isn't true. If you eat at a restaurant but forget your wallet, you can't pay your bill but you're still solvent. You have assets to cover your debts. "Can't pay your bill" is too broad a statement to be meaningful. There are many complicated financial instruments and needing time to make a payment doesn't automatically make you insolvent.


However, if SVB had EVEN longer, say, 10-15 years, they would have been fine, right?


> People keep talking about how 'this is a solvency crisis because if SVB had to sell everything today, they wouldn't cover liabilities' when that is the definition of a liquidity crisis.

I don't think that's right. It would be a liquidity crisis if the market value of everything they own is higher than their liabilities but they can't find a buyer at this time. You are saying that a liquidity crisis is when they can find a buyer but everything they have is worth less than their liabilities. That's not the case.


Over what time scale delineates the two? If they were say, 1000 year bonds, would you feel the same?

I don't know the answer, but I see both sides. As of today, their liabilities exceed their assets worth, so it does feel like solvency fits.


yep, it's a solvency issue. The minute they tried to offload the bonds at a price lower than they paid for them (which seemed like the right thing to do to fix the liquidity issue), they were effectively in a hole and even if they waited 10 years later, would not have been able to cover the deposits.

They NEEDED to raise capital.

Their liabilities > assets.

They were 100% insolvent.


You need to realize that essentially bond pricing reflects the present value of all cash flows you expect from the bond. For long term bonds especially, this makes it a particularly risky book, because you are very sensitive to interest rate changes.

If the interest rate goes up while you are holding your low interest bonds, that means that your future cash flow from the bond (the repayment) is literally worth less than what it was. Your bond payments have a lower real value due to the higher interest rate of the surrounding environment and the increasing price levels, despite being the same nominal amount. That's why the bond's price plummets in the market, which is why this is a solvency crisis: because the assets really are not good for the liabilities at present value, which is the only kind of valuation that makes sense here.


> The crux of the issue here is that, for many types of assets, banks are able to test whether they meet capital requirements based on the price they paid for the assets, rather than the price the assets are currently worth.

I think this will limit the types of assets banks can purchase. They'll need to purchase only assets that regularly trade (and thus are quoted) on the market.


The assets they were holding do regularly trade and could easily be valued. They knew that the value was down. The problem was they didn't actually have to do anything about it.


Because they planned to hold them to maturity, which meant that marking to market didn’t make much sense.

Of course here it blew up, but even marked to market assets will depreciate in value if a bank tries to liquidate tens of billions of dollars worth.


sounds like you are describing a liquidity crisis to me


A liquidity crisis is a solvency crisis if depositors are asking for their deposits. SVB made a miscalculation on their outflows and the price for that apparently is their whole market cap.


precisely, the maturation for the bonds were long-term. they would have been fine over time, if not for the bank run


If your "liquidity crisis" is measured in terms of years and years, you have a solvency crisis.

1 2023USD =/= 1 2033USD

Those aren't the same currency.

Solvency is whether you could go to market today (or soon) and show more assets than liabilities.


The liabilities are the same currency and the same amount in 2023 and 2033. They’re solvent.

The asterisk is that the liabilities are generating interest, but this is fine because (one very safely assumes) that’s covered by the interest on the long-duration assets SVB bought.


Interest rates have risen since those long-duration assets were purchased.

Depositors now expect higher interest rates which cannot be covered by those assets. They will withdraw their deposits and move them to a bank that can offer higher interest rates.


right, which is why this is a liquidity problem and not a solvency one. The assets cover the liabilities--you just have to wait ten years or whatever.


No.

Those are short term liabilities. The ability of depositors to withdraw their funds is part of the liability.

If the bank wanted to raise long term deposits it would have had to pay much higher interest rates.

Once they withdraw the bank can not raise additional funds because all other depositors have better interest rates in other banks.

No amount of time will cure this situation, the values of those bonds will not recover. This is not some temporary panic.

The only way for the situation to improve is if interest rates go down in the short term. Unless that happens, the bank remain insolvent.


No their bounds are worth fundamentally less than they paid for them.


There is a continuum in how quickly you need to sell and the price you willmget for something.


So it's the customer's fault?


No, it's Peter Thiel's fault for intentionally initiating a bank run?

SVB's systemic risk was just that the vast majority of their depositors are startups, and can corralled into action by VCs. Thiel decided all the startups he backed should withdraw all their money from SVB at once, and knew the rest of the VC world would follow suit.


> SVB's total equity was only $16bn

Which table in the 10-K would I look at to see this?


Agreed. Lots of people here in the comments are making assumptions about a system they don't understand. Depositors with > $250k aren't necessarily going to "take a haircut," for the reason you mentioned, plus a few others. Additionally:

1. Any financial advisor who recommended to these startups that they should keep >250k in a regular bank account should be fired. It's totally possible (and regularly done) to spread out cash among several financial institutions to protect against this very issue.

2. Any regular account with two or more signers (very typical for a business account) is insured up to 500k.

3. If spreading out your 6- or 7-figure assets to multiple institutions is too much of a burden, literally every business bank has special accounts or add-on features that either raise the FDIC default limit of 250k, or supplement it with external insurance. Again, if any startup's financial handlers didn't recommend this: fire them because they entirely failed to do their job.


SVB is not a typical regional bank taking deposits from middle-class workers, where most accounts are under the 250k insurance limit.

Banks that primarily serve ordinary workers typically have over 50% of total deposits in accounts that are under the limit, and are fully insured.

But for SVB, less than 3% of deposits are in accounts with less than $250k.

The cold, hard fact is that if the bank doesn't have sufficient assets to pay back depositors, no regulatory sleight of hand changes that fact.

There's a reason why large deposits aren't insured, and that's because the rich have the knowledge and resources to take care of themselves, and shouldn't co-opt the power of the state to force ordinary people to subsidize them when their bets go bad.

It would be hideously immoral to bail out fabulously wealthy VCs with funds from taxpayers and small depositors.


> The cold, hard fact is that the bank doesn't have sufficient assets to pay back depositors

This is not obvious at this time. This is not a cold hard fact. They absolutely, undeniably had assets in excess of depositor liabilities at the end of Dec 2022. They incurred losses since then. The accountants are still accounting.

> and no regulatory sleight of hand changes that fact.

There literally is regulatory "sleight of hand" to do this, and the FDIC has a demonstrated history of doing it, many times. When banks like this fail, they shop the assets and existing customers around to other banks, and critically: will pay the acquiring bank to close the gap between assets and liabilities + provide liquidity while assets mature.

> There's a reason why large deposits aren't insured, and that's because the rich have the knowledge and resources to take care of themselves, and shouldn't force ordinary people to subsidize them when their bets go bad.

The FDIC is entirely funded by insurance premiums paid for by the banks themselves. They are not tax payer funded.

Your fear is causing you to spread misinformation and scare people more than necessary.


> ... absolutely, undeniably ...

Were these assets marked to market, or were there billions in unrealized losses even then?

> regulatory "sleight of hand"

When claims on the FDIC exceed premiums paid, they are absolutely backed up by taxpayers.

There's a cap on what deposits are insured, it's not a secret. Large depositors knew their funds were not insured.

It's not "sleight of hand" to shop the assets in the market, maximize recovery, and make all depositors whole if possible. By all means do this!

But using FDIC funds to bail out uninsured depositors, or monkey business with quasi-government agencies backed by taxpayers paying above-market prices for securities, or agreeing to accept below-market interest rates on loans is definitely "sleight of hand" to obscure a direct subsidy to extraordinarily wealthy people.

By no means should insurance premiums paid on insured deposits be misappropriated to bail out uninsured depositors.


Being able to trust in the stability of banks is subsidizing the economy, not just wealthy individuals. It isn't CEOs who got tossed out of work in 2008, it was't wealthy individuals who had a 27% unemployment rate.

Wealthy individuals constantly try to game the system. Not letting the games they play hurt the rest of us is an entirely proper role for the government to take on.


I think the problem is when those games don't hurt the individuals playing them. After 2008, most of those banking executives who screwed things up saw few consequences.

I agree that it's a good move for the government to use taxpayer money to prevent an economic collapse. But that's still the lesser of two evils: the government should be working harder to disincentivize the kind of behaviors that make wealthy individuals think that playing these sorts of games is worth the risk.

I'm not sure what "working harder" should entail (I'm no expert on this sort of thing; others are), but I think it's pretty clear they're falling short.


> The cold, hard fact is that the bank doesn't have sufficient assets to pay back depositors.

That's not true. As of December 31, 2022, Silicon Valley Bank had approximately $209.0 billion in total assets and about $175.4 billion in total deposits. Even if liquidating assets forced a 15% haircut, depositors will be made whole.

If a bail out is necessary, it's likely in the form of a short-term loan to make depositors whole sooner rather than later. And, such a bail out can/should be structured as a loan with significant interest in which case it's not really a cost to taxpayers.


That $209 billion number was not marked to market. Their assets weren't worth that much then, and they're worth less now.

The cold, hard fact is that if the bank had sufficient assets to let depositors withdraw their cash, then they wouldn't be in receivership now.

You can argue that they actually do have sufficient assets, but they just can't access them right now, and we just need to wait ten years for bonds to mature.

But that argument only makes sense if you also say that large depositors should wait ten years to get their money out.

Naturally, a dollar that you might receive ten years from now is worth a lot less than a dollar you can actually spend today, and if regulatory sleight of hand obscures that essential fact to bail out the 1% of the 1%, that would be extraordinarily unjust.


Once again, my point is that your following statement is untrue: The cold, hard fact is that the bank doesn't have sufficient assets to pay back depositors.

It is not a cold, hard fact that they don't have sufficient assets. It's entirely possible that after liquidating their assets (on the time scale of months), they can make depositors whole even after factoring in the time cost of money.


Of course you're correct, nobody knows the future.

It's possible the $80 billion in mortgage-backed securities they bought yielding 1% and maturing in 10 years might be salvaged. They're now worth far less than they paid because interest rates are up to 5% and nobody wants to buy bonds that yield 1%.

Maybe this receivership will bork a thousand startups, throw a bunch more people out of work, and cascade into a larger recession, driving interest rates back down to 1%, and restoring the value of the mortgage-backed securities, and making the bank solvent again!

Or maybe a recession severe enough to drive interest rates back to 1% would also be severe enough to prevent homeowners from paying their mortgages, which wouldn't be great for the valuation of mortgage-backed securities.


Not unless interest rates plummet, no. The bulk of portfolio losses is from long term bonds losing value with the increase in interest rates. Allowing more time to sell doesn’t help you there, especially if you have to pay prevailing interest rates on any delay.


You’re basically saying the employees at a non-tribal number of startups should not be paid payroll because “fat cat VCs”.


No, I'm saying let the fat cat VCs fund the payroll, not a quasi-governmental fund backed by taxpayers that explicitly said uninsured deposits are not insured.


> 3. If spreading out your 6- or 7-figure assets to multiple institutions is too much of a burden, literally every business bank has special accounts or add-on features that either raise the FDIC default limit of 250k, or supplement it with external insurance. Again, if any startup's financial handlers didn't recommend this: fire them because they entirely failed to do their job.

Contractual obligations often prevent this, btw. Many SVB customers had loans with SVB, which prevented them from using other banks.


Wow, is this common? From a systems perspective it seems like pure folly, increasing systemic risk and reducing resilience. (from SVB’s perspective I’m sure it seemed great …) it feels like it should be illegal !


I asked a handful of people with experience in those things and they all said it is very common. I was very surprised as well.


I don't think the signers count towards the limit like a personal joint account does and would keep a business account at $250k.

https://www.usbank.com/bank-accounts/fdic-deposit-insurance-...


For Pete's sake, 1 month Treasury bills are yielding more than almost every bank account. Just set up a TreasuryDirect account and be done with it.


> 2. Any regular account with two or more signers (very typical for a business account) is insured up to 500k.

Is it typical for business accounts to be owned by someone other than the business?


To be clear, if you have 10 million cash, and you need >250k to make payroll a couple of times while your banks assets are sold off and you get the rest of your cash back (or 95%, or whatever), there is no need to spread that 10 mil over 40 different banks. That would be very inconvenient. Just to set it up so you have enough insurance to help you survive a short while. You can put 9.75M in one account and 250k in another bank.


How does that work with payroll? 500k is not enough for a moderately sized startup to make payroll.


> Any financial advisor who recommended to these startups that they should keep >250k in a regular bank account should be fired.

Looks like not just startups, e.g. Roku had ~500M in SVB

https://edition.cnn.com/2023/03/10/business/roku-svb-cash/in...


It seems like SVB was exploiting the lack of sophistication among early startups, and now it is going to bite all those startups in the ass.

Wealthfront offers me as an individual better tools for managing risk that it sounds like people who banked with SVB were getting.


The thing that's strange is FDIC took control and setup a receiving bank for liquidation.

That's not normal; FDIC works quite hard to find a bank willing to take over - usually they can work out what the "cost" is to take over, and FDIC pays the receiving bank that amount to "eat" the dying one.

If they don't announce they have a bank to assume SVP by Monday, it's quite abnormal.


Exactly. The fact they didn't have a bank lined up points very heavily to the fact that they are not going to be made completely whole. In the past, the FDIC has found a buyer and as part of that process guarantees some amount of the losses. IE: Savior Bank buys Failed Bank for pennies on the dollar, or even for a negative amount. They get all deposits, insured or not, and all assets--meaning loans. Then, the FDIC guarantees they'll make Savior Bank whole some percentage of losses on assets that go bad. Sometimes 80% or more

This arrangement usually results in all depositors being made whole, and the FDIC fund not taking any, or many, actually losses, because most of the loans will still pay back, at least partially

They also do this before seizing the bank, so that it's all very orderly and calms any panics

SVB was looking for a buyer on the open market and that failed--no surprise

That the FDIC also could not find a buyer that they could subsidize and announce the same day as the seizure is very damning. I would be shocked if someone comes in later and tries to buy it

Also, the fact the seizure happened in the middle of the business day, and not at the close of business yesterday points to a somewhat disorderly and rapidly deteriorating condition. I think maybe the run picked up a lot faster than they expected


That's a really good point, FDIC usually swoops in Friday night; and this was closed on a Friday during business hours, that's actually insane; they couldn't hold on 8-10 more hours, the run must have been really bad.

Even WAMU (a huge FDIC action) was done after banking hours closed on the west coast. https://en.wikipedia.org/wiki/Washington_Mutual


It doesn't really work this way. It takes time to set up acquisitions, even by the FDIC, and the "FDIC stormtroopers sell bank overnight" is apocryphal. They do weeks of legwork leading up to the actual handover of the bank.

We don't know the timeline here, but speculating that "it must be bad" because things didn't happen overnight isn't really responsible.


During 2008 crisis it worked exactly as described. Most failures resulted in the failing bank’s acquisition being decided before the seizure happened. The way this happened is very rare


I think those actions worked out of hours because the bank's bosses let the FDIC know they were insolvent before a run happened, though? This time the bosses thought they could hold on.

If so, then the difference doesn't imply much about the banks asset level, just the idiocy of its bosses.


I mean...I think we're describing the same thing? So sure, it worked "exactly as described", but I think most folks on HN have no idea what "as described" means, in this context.

A lot of folks (like the top of thread) want to suggest that because it didn't play out like the 60 minutes story, it must mean something significant. That story barely touched on the weeks of leg-work the FDIC did for that particular bank, for example. If you weren't paying close attention, you'd miss it.


I just saw recent footage of 40-50 people standing in line outside an SVB branch waiting for service (withdrawals most likely).

I would surmise that the news spread fast among the tech demographic the bank services and the run is really bad.


And in this case, tech savy individuals are willing to send $XXM in 6 clicks, 3min after getting a slack message. That's a pretty quick kind of run relative to old-school 'stand in line for your personal life savings' kind of run.


It seems like there’s a lot of uncertainty around the dollar amount of the deposits in excess of FDIC limits which would make it difficult to figure out a deal.


Most FDIC bank takeovers are slow moving crashes, allowing for a longer negotiation where buyers can evaluate the loan portfolio they are buying. This is a reaction to a classic run, so no time for that.


So what I think is happening here is that if you take over a bank the traditional way, you need to mark-to-market all of the bank's assets -- so all of the losses from the long-dated MBS that would be perfectly fine if held to maturity would have to be recognized immediately, just crushing the balance sheet of anyone who bought it. Probably trying to line someone up who can either absorb that loss, figure out a way to recapitalize without recognizing the loss, or get access to some other lending facility.


But from what I read it's common for FDIC to "sell" to the acquiring bank at a price that wouldn't make a loss.

So if you marked to market all those long term bonds and then sold SVB to a bigger bank at the resulting (possibly negative) valuation. Why wouldn't a big bank take that deal?


The size of this failed bank is quite abnormal. The business of this failed bank is quite abnormal. The surprising thing may be that the find anyone willing to take it out of their hands - let alone by Monday.


When was the last US bank failure where depositors lost money?


Appears to be 2020, from a cursory Google search.

https://money.stackexchange.com/questions/129772/has-anyone-...

The oft quoted 'nobody has lost money' is from the FDIC, and specifically says 'insured deposits.'


Chase has wanted this bank for years. Either they are taking advantage (takeover) or <conspiracy theory unfounded> helped accelerate it.


Most of the good staff were poached by First Republic over the past few years. That caused me to bank my current startup at FR after more than 25 years of SVB.

So I don’t think there was much for Chase to buy. SVB has been in decline for a while


Just saw a First Republic stock chart in crash territory, contagion?


S&P bank sector is down across the board, but presume it will regress to the mean fairly soon. Aren't many other banks with exposure similar to SVB's.


First republic is in the same region and has a ton of tech clients. Collateral damage. So far it looks like they don't have the same exposure as SVB did though.



Interesting you say this, their stock is up despite rest of my portfolio being red


Flight to safety? Like when there was the Southwest Airlines December debacle - AAL skyrocketed for a bit.


What is "SVP"?


I suspect a typo for SVB, "Silicon Valley Bank".


93% of deposits are uninsured according to a recent regulatory filing. So, I doubt it will be “all depositors” that would be made whole.

https://twitter.com/business/status/1634211584657571843?s=20


FDIC means you get paid immediately regardless, but the bank will almost invariably turn out to be good for the rest, just not quickly.


No, FDIC means you get paid up to $250,000 per insured account immediately Any amount over that you are not guaranteed to be able to withdraw. These people are a long way from regardless


> No, FDIC means you get paid up to $250,000 per insured account immediately Any amount over that you are not guaranteed to be able to withdraw.

The first half is what FDIC insurance means. The FDIC helps distressed banks in a variety of ways, including insuring deposits under $250K, but typically all depositors are made mostly whole again.


IndyMac depositors got 50 cents on the dollar above the insured amount in the 2008 collapse.

https://www.ocregister.com/2009/03/20/depositor-losses-final...


Typically the FDIC finds another institution to buy the distressed bank, and backstops losses on bad assets. They were not able to do that in this case

I suspect many depositors will be taking losses in this failure


> but typically all depositors are made mostly whole again

Typically, but no guarantee. And on what time frame, nobody knows.