In the wake of Silicon Valley Bank's failure, many financial journalists reveal they don't understand banks. Do you? Let's take advantage of a "learning moment."
Thanks for a clear, simplified explanation of today’s financial system. That we skate on very thin and truly mythical ice is lost on far too many innocent people. That our financial system is corrupted is something I’m well aware of yet the wealth of the nation is not in the bank deposits though that productive wealth can easily be destroyed by mismanagement in the financial system. Hope you are able to see a way out of the mess we are in. I personally think only by removing government control from the money system could we ever have financial stability. I will read more of what you have written as this is the first post of yours that I have read.
The essence of SVB as I see it discussed is that enough people/companies wanted their $$ out of SVB at the same time, and SVB didn't have sufficient $$ to cover it. Is this description accurate? (What confuses me is squaring it with your real bank chart, which seems to suggest that every dollar put into SVB remained in SVB, as the $100 stays constant on your chart.)
Every dollar deposited in a bank (an entry on the liability side) carries with it corresponding "reserves" on the asset side. If deposits leave the bank to another bank, an equal amount of reserves goes with them. So reserves come and go when already-existing deposits are transfered into or out of a bank.
No reserves do not leave the bank when the bank invests. When the bank invests, it creates new money in the form of bank deposits to fund the investment. The Real Bank chart doesn't "seem to suggest" that reserves do not change when the bank invests. It is saying precisely that.
The essence of a bank run (not the essence of banking) is that more people want their money out than there are liquid assets to fund the withdrawals. But as the article points out, this can happen in other financial institutions, not just banks.
Best to explain it this way: Look at "Real Bank" on my Substack. If the bond investment declines in value by $10, then total assets go down to $170. But there are still $180 in demand deposits. So then if all the depositors want their money, the bank can sell every asset and turn it to cash and still fall short by $10, so it is bankrupt.
In pattern, that is what happened to SVB. One way to avoid this is to keep reserve cash levels stable by insuring your deposits are stable. (When deposits flow out, reserves flow out.) An example of stable deposits would be business deposits that go up when company revenues come in and go down when payroll is made, but always maintain a predictable minimum deposit level. My view of SVB is they allowed too many unstable deposits into their bank, which they thought were "sticky" or stable. They didn't understand their deposit base. They had "toxic deposits." Does that explanation help?
Thank you so much! I'm still confused, but that's on me I think. Is it because when the depositor closes out their account then everything goes away? Zeros out? And as that keeps happening all the actual money and all of the created money goes poof?
Suppose the depositor moves his money to another bank, or suppose he writes a check to someone who deposits that check in another bank. In that event, both deposits and reserves move to the other bank in equal amounts. So, looking at Real Bank, if the customer writes a check for 10 so someone, when the receiver of the check deposits it in his own bank, he gets a deposit of 10 in his bank, AND i10 in cash reserves moves to the new bank. So, back in Real Bank, that would leave 90 in cash reserves and 80 in investments on the asset side and 170 in deposits on the liability side. But no money goes away, it just moves to another bank. If the depositor takes his money out in cash, it is basically the same for Real Bank: 170 left in deposits, and 90 left in reserves. The 10 taken out is now paper cash and becomes a direct liability of the central bank. The 10 cash is not in the banking system, but it is still included in the money supply, which consists of bank deposits plus currency in circulation.
Phew!! Thank you! I am digesting. The whole assets/liabilities thing is confusing me. Are they "liabilities" because the depositor can ask for them and the bank has to fork them over? (Sorry to be so slow about these things. Thank you again for answering... do not hesitate to stop answering.)
Every dollar deposited in a bank (an entry on the liability side) carries with it corresponding "reserves" on the asset side. If deposits leave the bank to another bank, an equal amount of reserves goes with them. So reserves come and go when already-existing deposits are transfered into or out of a bank.
No reserves do not leave the bank when the bank invests. When the bank invests, it creates new money in the form of bank deposits to fund the investment. The Real Bank chart doesn't "seem to suggest" that reserves do not change when the bank invests. It is saying precisely that.
The essence of a bank run (not the essence of banking) is that more people want their money out than there are liquid assets to fund the withdrawals. But as the article points out, this can happen in other financial institutions, not just banks.
What financial institutions operate in this way other than banks?: Take deposits, promise to pay a certain amount, invest money to make profit on the difference between return on investment and the interest payment (if any) on deposit?
My question comes from the inability to think of any other institution like that. Whether it's a mutual fund, an insurance company, a hedge fund, or any other, none of them promise you a certain amount. What you get is the value of assets at the time of withdrawal (and hopefully you come in to those institutions knowing this fact). So when you stress this distinction between "runs on financial intermediaries" and "runs on banks as money creators", what institutions do you have in mind?
You mentioned money market fund, but they don't guarantee your principal as banks do in their promise of "payment on demand", do they? That's the essential difference in my eyes.
Hey Robertas. Best example is a money market fund (which is a money market MUTUAL fund). which I mentioned in the essay and podcast You give the fund money. They give you shares valued at $1 each. They invest the money in short term bonds (<than one year maturity). They try (usually succeed) to permit redemption any time at one dollar per share. The Fund pays you interest in the form of new shares: they can always try to peg the share value at $1 per share. to make their product more money-ish. Under extreme conditions - either credit problems in the fund's investments, or even rapidly rising interest rates AND a simultaneous demand to redeem shares for cash - a money market fund can fail to have the liquidity to pay you out at $1 per share. This is called "breaking the buck" in money market language. It would have happened in the GFC, but the Fed stepped in and saved the private commercial paper market from default. What I describe is less of a problem today, because 75% of money market fund assets are invested in US Treasurys or Fed reverse repos, which are obviously at very low risk of default. But it is still possible if everyone wants dollars at the same time. Other financial institutions, like insurance companies, do not allow full withdrawals on daily demand, but they do guarantee withdrawals of specific amounts at specific times. They can fail to pay depositors due to asset/liability mismatch. So in fact a money market fund DOES promise you a specific amount at a specific time. So does a life insurance company depending on the contract you sign. So does a defined-obligation pension fund. It's just that their obligations are usually not payback in full,"on demand." Another recent example is some shares in REITs done by Blackrock that promise on-demand redemptions at NAV. Yes, the NAV changes with underlying prices (as determined by BR) but even they recently been unable to liquidate at their posted NAV, so have delayed payments.
Also, 2nd related question. Your examples assume that when banks make loans or buy bonds, they create deposits on the liability side. But that only makes sense if the money paid to the previous bond holder/borrower stays in the bank, which is rarely true, isn't it? A loan is taken to pay somebody else, and thus the deposit is withdrawn and can be dispersed among multiple banks of those who get paid. So the balance sheet immediately looks different - as people withdraw cash from loans, the cash reserves should go down, shouldn't they?
I think it's best to answer you line by line, in caps: "Also, 2nd related question. Your examples assume that when banks make loans or buy bonds, they create deposits on the liability side. [CORRECT ] But that only makes sense if the money paid to the previous bond holder/borrower stays in the bank, which is rarely true, isn't it? [THE NEWLY CREATED MONEY PAID TO THE BOND OWNER CAN AND OFTEN DOES MOVE TO ANOTHER BANK - BUT IT IS STILL NEWLY CREATED MONEY] A loan is taken to pay somebody else, and thus the deposit is withdrawn and can be dispersed among multiple banks of those who get paid. [SURE, YOU CAN MOVE YOUR DEPOSIT TO ANY BANK ACCOUNT AT ANY BANK UNDER FED SUPERVISION. WHEN YOU USE THAT NEW MONEY TO PAY SOMEONE ELSE, IT JUST MOVES TO THEIR BANK ACCOUNT, LIKELY AT SOME OTHER BANK] So the balance sheet immediately looks different - as people withdraw cash from loans, the cash reserves should go down, shouldn't they? [YES, CASH RESERVES GO DOWN DOLLAR FOR DOLLAR AT THE ORIGINATING BANK AS DEPOSITS ARE WITHDRAWN. BUT THESE RESERVES "TRAVEL" WITH THE DEPOSITS TO OTHER BANKS. SO IN CREATING MONEY, THE CREATING BANK ACQUIRES BOTH A NEW ASSET (LOAN OR BOND) AND A NEW LIABLITY (DEPOSIT). CASH RESERVES STAY THE SAME WHEN MONEY IS CREATED. WHEN THE NEW DEPOSIT MOVES TO ANOTHER BANK, THE NEW ASSET STAYS WITH THE ORIGINAL BANK, WHILE BOTH DEPOSITS AND CASH RESERVES MOVE TO THE OTHER BANK. SO THE NEW DEPOSIT MONEY STAYS IN THE BANKING SYSTEM, JUST BECOMES AN OBLIGATION OF A DIFFERENT BANK, BACKED BY THE RESERVES THAT MOVED FROM THE FIRST BANK TO THE SECOND BANK. I HOPE THAT HELPS.
Hey Gary. Government spending generates bank deposits. When government (the Treasury) sends a social security check or a check to a government contractor, these checks are deposited into the receiver's bank account. Deficit spending just means the government spent more than it collected in taxes, so no matter where the money came from (taxes or borrowing) when Treasury sends out a check it will increase bank deposits..
But here is an interesting point. Government collection of revenue decreases bank deposits. When you pay taxes, deposits leave the bank and go to Treasury, who deposits your money in its account at the Fed. That means the money supply temporarily decreases, because Treasury's general account at the Fed is not counted in the money supply (except by some Austrians). But Treasury spends that money pretty quickly, so then it goes right back into bank deposits.
When Treasury borrows from you (say you buy a bond direct from the Treasury) the same thing happens: Treasury gets your money, it leaves your bank, and is held in Treasury's account at the Fed, and the money supply (bank deposits) decreases. But again, not for long, because Treasury spends it pretty fast, so it re-enters the banking system.
When the accumulated deficit (total debt) gets large enough, central banks monetize some of that debt via some program like QE. In that scenario, the central bank's purchase of the government's debt does generate deposits. So deficits do generate deposits in the long run in this roundabout way. But taxing and borrowing and spending, without monetization, does not increase deposits on a net basis, just moves the money around. Sorry for the long answer.
The dialogue with Gary was helpful for me. It seems the government can raise interest rates as it is doing now (and therefore reduce private party borrowing and inflation) but nonetheless print money and spend it (in a variety of ways) such that each existing dollar in existence is devalued. Do I understand this correctly? If so, then simultaneously the government can raise interest rates to battle inflation but also print money to make my dollars saved increasingly worth less. Do I have this right?
Right. Government creates new money by "monetizing" their own debt, which is the process of buying their own bonds with newly created money (as in QE). Government can do this whether interest rates are high or low. "Monetizing" tends to drive interest rates lower, by creating a demand for government bonds that didn't exist in the market. But when inflation (government money creation) gets to an extreme, no amount of monetizing can overcome the inflation component of the interest rate. That's the point when the bond market gives up on government debt, just refuses to buy it. We are not there yet.
Thanks for a clear, simplified explanation of today’s financial system. That we skate on very thin and truly mythical ice is lost on far too many innocent people. That our financial system is corrupted is something I’m well aware of yet the wealth of the nation is not in the bank deposits though that productive wealth can easily be destroyed by mismanagement in the financial system. Hope you are able to see a way out of the mess we are in. I personally think only by removing government control from the money system could we ever have financial stability. I will read more of what you have written as this is the first post of yours that I have read.
Thanks for reading and hanks for the comment.
The essence of SVB as I see it discussed is that enough people/companies wanted their $$ out of SVB at the same time, and SVB didn't have sufficient $$ to cover it. Is this description accurate? (What confuses me is squaring it with your real bank chart, which seems to suggest that every dollar put into SVB remained in SVB, as the $100 stays constant on your chart.)
Every dollar deposited in a bank (an entry on the liability side) carries with it corresponding "reserves" on the asset side. If deposits leave the bank to another bank, an equal amount of reserves goes with them. So reserves come and go when already-existing deposits are transfered into or out of a bank.
No reserves do not leave the bank when the bank invests. When the bank invests, it creates new money in the form of bank deposits to fund the investment. The Real Bank chart doesn't "seem to suggest" that reserves do not change when the bank invests. It is saying precisely that.
The essence of a bank run (not the essence of banking) is that more people want their money out than there are liquid assets to fund the withdrawals. But as the article points out, this can happen in other financial institutions, not just banks.
So Jim, in the case of SVB then, was the money still "there" even after it had been invested, and if it was, then why did SVB fail?
Best to explain it this way: Look at "Real Bank" on my Substack. If the bond investment declines in value by $10, then total assets go down to $170. But there are still $180 in demand deposits. So then if all the depositors want their money, the bank can sell every asset and turn it to cash and still fall short by $10, so it is bankrupt.
In pattern, that is what happened to SVB. One way to avoid this is to keep reserve cash levels stable by insuring your deposits are stable. (When deposits flow out, reserves flow out.) An example of stable deposits would be business deposits that go up when company revenues come in and go down when payroll is made, but always maintain a predictable minimum deposit level. My view of SVB is they allowed too many unstable deposits into their bank, which they thought were "sticky" or stable. They didn't understand their deposit base. They had "toxic deposits." Does that explanation help?
Thank you so much! I'm still confused, but that's on me I think. Is it because when the depositor closes out their account then everything goes away? Zeros out? And as that keeps happening all the actual money and all of the created money goes poof?
Suppose the depositor moves his money to another bank, or suppose he writes a check to someone who deposits that check in another bank. In that event, both deposits and reserves move to the other bank in equal amounts. So, looking at Real Bank, if the customer writes a check for 10 so someone, when the receiver of the check deposits it in his own bank, he gets a deposit of 10 in his bank, AND i10 in cash reserves moves to the new bank. So, back in Real Bank, that would leave 90 in cash reserves and 80 in investments on the asset side and 170 in deposits on the liability side. But no money goes away, it just moves to another bank. If the depositor takes his money out in cash, it is basically the same for Real Bank: 170 left in deposits, and 90 left in reserves. The 10 taken out is now paper cash and becomes a direct liability of the central bank. The 10 cash is not in the banking system, but it is still included in the money supply, which consists of bank deposits plus currency in circulation.
Phew!! Thank you! I am digesting. The whole assets/liabilities thing is confusing me. Are they "liabilities" because the depositor can ask for them and the bank has to fork them over? (Sorry to be so slow about these things. Thank you again for answering... do not hesitate to stop answering.)
Thanks for reading and thanks for the comment.
Every dollar deposited in a bank (an entry on the liability side) carries with it corresponding "reserves" on the asset side. If deposits leave the bank to another bank, an equal amount of reserves goes with them. So reserves come and go when already-existing deposits are transfered into or out of a bank.
No reserves do not leave the bank when the bank invests. When the bank invests, it creates new money in the form of bank deposits to fund the investment. The Real Bank chart doesn't "seem to suggest" that reserves do not change when the bank invests. It is saying precisely that.
The essence of a bank run (not the essence of banking) is that more people want their money out than there are liquid assets to fund the withdrawals. But as the article points out, this can happen in other financial institutions, not just banks.
Hi Jim.
What financial institutions operate in this way other than banks?: Take deposits, promise to pay a certain amount, invest money to make profit on the difference between return on investment and the interest payment (if any) on deposit?
My question comes from the inability to think of any other institution like that. Whether it's a mutual fund, an insurance company, a hedge fund, or any other, none of them promise you a certain amount. What you get is the value of assets at the time of withdrawal (and hopefully you come in to those institutions knowing this fact). So when you stress this distinction between "runs on financial intermediaries" and "runs on banks as money creators", what institutions do you have in mind?
You mentioned money market fund, but they don't guarantee your principal as banks do in their promise of "payment on demand", do they? That's the essential difference in my eyes.
Hey Robertas. Best example is a money market fund (which is a money market MUTUAL fund). which I mentioned in the essay and podcast You give the fund money. They give you shares valued at $1 each. They invest the money in short term bonds (<than one year maturity). They try (usually succeed) to permit redemption any time at one dollar per share. The Fund pays you interest in the form of new shares: they can always try to peg the share value at $1 per share. to make their product more money-ish. Under extreme conditions - either credit problems in the fund's investments, or even rapidly rising interest rates AND a simultaneous demand to redeem shares for cash - a money market fund can fail to have the liquidity to pay you out at $1 per share. This is called "breaking the buck" in money market language. It would have happened in the GFC, but the Fed stepped in and saved the private commercial paper market from default. What I describe is less of a problem today, because 75% of money market fund assets are invested in US Treasurys or Fed reverse repos, which are obviously at very low risk of default. But it is still possible if everyone wants dollars at the same time. Other financial institutions, like insurance companies, do not allow full withdrawals on daily demand, but they do guarantee withdrawals of specific amounts at specific times. They can fail to pay depositors due to asset/liability mismatch. So in fact a money market fund DOES promise you a specific amount at a specific time. So does a life insurance company depending on the contract you sign. So does a defined-obligation pension fund. It's just that their obligations are usually not payback in full,"on demand." Another recent example is some shares in REITs done by Blackrock that promise on-demand redemptions at NAV. Yes, the NAV changes with underlying prices (as determined by BR) but even they recently been unable to liquidate at their posted NAV, so have delayed payments.
Also, 2nd related question. Your examples assume that when banks make loans or buy bonds, they create deposits on the liability side. But that only makes sense if the money paid to the previous bond holder/borrower stays in the bank, which is rarely true, isn't it? A loan is taken to pay somebody else, and thus the deposit is withdrawn and can be dispersed among multiple banks of those who get paid. So the balance sheet immediately looks different - as people withdraw cash from loans, the cash reserves should go down, shouldn't they?
I think it's best to answer you line by line, in caps: "Also, 2nd related question. Your examples assume that when banks make loans or buy bonds, they create deposits on the liability side. [CORRECT ] But that only makes sense if the money paid to the previous bond holder/borrower stays in the bank, which is rarely true, isn't it? [THE NEWLY CREATED MONEY PAID TO THE BOND OWNER CAN AND OFTEN DOES MOVE TO ANOTHER BANK - BUT IT IS STILL NEWLY CREATED MONEY] A loan is taken to pay somebody else, and thus the deposit is withdrawn and can be dispersed among multiple banks of those who get paid. [SURE, YOU CAN MOVE YOUR DEPOSIT TO ANY BANK ACCOUNT AT ANY BANK UNDER FED SUPERVISION. WHEN YOU USE THAT NEW MONEY TO PAY SOMEONE ELSE, IT JUST MOVES TO THEIR BANK ACCOUNT, LIKELY AT SOME OTHER BANK] So the balance sheet immediately looks different - as people withdraw cash from loans, the cash reserves should go down, shouldn't they? [YES, CASH RESERVES GO DOWN DOLLAR FOR DOLLAR AT THE ORIGINATING BANK AS DEPOSITS ARE WITHDRAWN. BUT THESE RESERVES "TRAVEL" WITH THE DEPOSITS TO OTHER BANKS. SO IN CREATING MONEY, THE CREATING BANK ACQUIRES BOTH A NEW ASSET (LOAN OR BOND) AND A NEW LIABLITY (DEPOSIT). CASH RESERVES STAY THE SAME WHEN MONEY IS CREATED. WHEN THE NEW DEPOSIT MOVES TO ANOTHER BANK, THE NEW ASSET STAYS WITH THE ORIGINAL BANK, WHILE BOTH DEPOSITS AND CASH RESERVES MOVE TO THE OTHER BANK. SO THE NEW DEPOSIT MONEY STAYS IN THE BANKING SYSTEM, JUST BECOMES AN OBLIGATION OF A DIFFERENT BANK, BACKED BY THE RESERVES THAT MOVED FROM THE FIRST BANK TO THE SECOND BANK. I HOPE THAT HELPS.
Hi Jim, thank you for the article. Can I ask, do government deficits generate deposits as well?
Hey Gary. Government spending generates bank deposits. When government (the Treasury) sends a social security check or a check to a government contractor, these checks are deposited into the receiver's bank account. Deficit spending just means the government spent more than it collected in taxes, so no matter where the money came from (taxes or borrowing) when Treasury sends out a check it will increase bank deposits..
But here is an interesting point. Government collection of revenue decreases bank deposits. When you pay taxes, deposits leave the bank and go to Treasury, who deposits your money in its account at the Fed. That means the money supply temporarily decreases, because Treasury's general account at the Fed is not counted in the money supply (except by some Austrians). But Treasury spends that money pretty quickly, so then it goes right back into bank deposits.
When Treasury borrows from you (say you buy a bond direct from the Treasury) the same thing happens: Treasury gets your money, it leaves your bank, and is held in Treasury's account at the Fed, and the money supply (bank deposits) decreases. But again, not for long, because Treasury spends it pretty fast, so it re-enters the banking system.
When the accumulated deficit (total debt) gets large enough, central banks monetize some of that debt via some program like QE. In that scenario, the central bank's purchase of the government's debt does generate deposits. So deficits do generate deposits in the long run in this roundabout way. But taxing and borrowing and spending, without monetization, does not increase deposits on a net basis, just moves the money around. Sorry for the long answer.
The dialogue with Gary was helpful for me. It seems the government can raise interest rates as it is doing now (and therefore reduce private party borrowing and inflation) but nonetheless print money and spend it (in a variety of ways) such that each existing dollar in existence is devalued. Do I understand this correctly? If so, then simultaneously the government can raise interest rates to battle inflation but also print money to make my dollars saved increasingly worth less. Do I have this right?
Right. Government creates new money by "monetizing" their own debt, which is the process of buying their own bonds with newly created money (as in QE). Government can do this whether interest rates are high or low. "Monetizing" tends to drive interest rates lower, by creating a demand for government bonds that didn't exist in the market. But when inflation (government money creation) gets to an extreme, no amount of monetizing can overcome the inflation component of the interest rate. That's the point when the bond market gives up on government debt, just refuses to buy it. We are not there yet.
Extremely clear answer, thank you..and now I see the link between QE and deficits in growing deposits.