Post comments on any MMT related topic. Civil discussion and no coarse language. Keep all discussions threaded (use the reply button on the comment you are replying to) where possible please.
There are a number of approaches/”schools” in the field of money/monetary systems and a number in actual practice. Regarding the somewhat hybrid system used in the USA, I would be very interested in an MMT discourse on two of the major recommendations for change of the US system, namely Ellen Brown’s public bank addition to the system and second, the changes described in http://www.realmoneyecon.com. Comparisons with the MMT approach would, of course, be of great interest.
I came to MMT after reading Ellen Brown’s “The Web of Debt”. She was mainly concerned that money was all debt, and to her that implied that it was a Ponzi scheme, because not enough money was created by lending to cover the interest required on the debt. At some point a debtor would not be able to find money to pay the interest on his/her loan and would default through no fault of their own.
Her book is an excellent coverage of the history of fiat money, and she is committed to it, but is discouraged, since the Great Recession did not lead to some basic changes. She decided to emphasize an alternative approach to totally changing the monetary system of the US: public banks owned and regulated by public entities. These could be local, state, city owned banks. They would not engage in derivatives, CDO’s and other shady enterprises but be plain vanilla commercial savings and loans banks owned by these communities. She is currently concerned with efforts to sneak in wording in the TransPacificTrading Treaty that only private banks be authorized. Those would trump the US Constitution as a Treaty. But it would also make impossible the public banks she is working to get established.
The “Real Money Movement” seems to be a British/European moviment with ideas very similar to those of MMT. I’ve read WHERE DOES MONEY COME FROM? by Ryan-Collins, Josh; Tony Greenham; Richard Werner; Andrew Jackson (2013-09-24). Where Does Money Come From? . The New Economics Foundation. Kindle Edition. They favor fiat money. I don’t know much more than that at the moment.
So, there are groups with similar fiat-money anti-austerian positions. What would unite them politically, I don’t know.
All money is basically fiat money. Pre 1933 in the USA we had “gold backing” but you had to have “fiat” or faith that the banks had the gold to back the gold certificates which they did not have. Originally, under law, banks were required to have only 25% backing with gold. Fractional reserve banking allows banks to create money. EU banks currently operate on a 1%reserve requirement. They are really greedy. The current rule in the USA is somewhere in the 8%range
There are various proposals being put forward that would require a change in the current rules, which is almost impossible to do politically. It’s unlikely to happen anytime soon. People are welcome to float proposals, but they also need to think about how this could be implemented practically. It would require garnering powerful support.
MMT founder Warren Mosler has proposed the simple alternative of the Fed setting the policy rate permanently to zero and providing solvent member institutions unlimited liquidity, with the Treasury not issue securities of duration longer than 3 months, which make T-bills effectively cash equivalents. The Fed chair and Treasury secretary could do this immediately simply by altering current policy.
This could be done within the current rules just by tweaking policy. It would accomplish what the other proposals are attempting to do, other than proposals that would eliminate banks from money creation and make them pure intermediaries, lending out savings.
Some MMT proponents favor dual sources of money creation as now — government currency issuance (settlement balances are cash equivalents being exchangeable at par for cash at the central banks) and bank credit that is subordinate since it clears in currency.
Bill Mitchell recommends full bank nationalization on the other hand, treating money creation as a public utility.
I like Ellen Brown’s back door approach and I like the frontal assault described in http://www.realmoneyecon.com “They” are being surrounded with Mosler charging on the flank!
This may be a stupid question, but I am struggling to understand why it is when “creating money out of thin air ” is
a) Good and/or
Is it solely dependent on who is doing it?
a) the country’s Central Bank or b) Commercial Banks
Before even learning about MMT, I had come to the conclusion that a government must print & issue it’s country’s money as a responsibility to ensure the proper functioning of its economy.
Then, after reading a few books on the subject of the world’s finanacial woes, I was being told that printing money out of nothing was “bad” for a country’s economy.
So what is the correct answer? I am confused!
Yes, “who is doing it” is very important and who enjoys the benefits of the seigniorage is the real issue. We have a hybrid system. The government mints coins and spends them at face value. Pennies and nickels have negative seigniorage with the other coins have increasing seigniorage as the value increases. The proposed T$ coin would have carried a very large seigniorage that would have benefited the citizens of the nation because it would have meant more services without more taxes. The primary money used in the economy is not really money per se but it used as money and is more properly referred to as credit. Banks have an overhead expense but money creation is essentially free to them. They can loan $1000 or tap the “0” key three more times to loan 1 M and it doesn’t cost them any more. Commercial banks, privately owned entities, then collect interest on the seigniorage of the credit money they put into the economy. Many people believe all benefits of seigniorage should accrue to the citizens of the nation and not to privately owned entities. Proposals to change the present or to force it in a direction to enable the citizens to enjoy the benefits of seigniorage All such proposals must retain the benefits and necessity of credit in the economy. The two most serious proposals for change are described in http://www.realmoneyecon.org which advocates ending fractional reserve banking a la the Chicago Plan from the 1930s. The other is Ellen Brown’s public bank initiative described in her Web of Debt blog and in her books. Minting T$ coins is a back door approach that would eventually accomplish the same end.
It is neither good nor bad. It depends on how, when & where the money circulates and if products and services are produced and how that affects the overall economy. This in the end determines whether it is bad or not. It may in turn be affected by your morals/ethics/politics whether it is good or bad as well.
Tom Hickey recommended I read Can Taxes and Bonds Finance Government Spending? by Stephanie Bell [Kelton now (1998). Prof. Stephanie Kelton I admire very much for her efforts to propagate MMT and her excellent talks to lay groups about MMT. Her article is mostly about problems with reconciling inflows with outflows in the banking system.
However, I think she committed a fallacy when she argued on p. 20-22:
“Fortunately, there is another, more powerful, method by which to argue that taxation and bond sales should not be considered financing operations. The argument is a technical one and requires an understanding that Federal Reserve Notes (and reserves) are booked as liabilities on the Fed’s balance sheet and that these liabilities are extinguished/discharged when they are offered in payment to the State. It must be recognized that when currency reserves return to the State, the liabilities of the State are reduced and high-powered money is destroyed.”
“The destruction of these promises is no different from the private destruction of a promise once it has been fulfilled. In other words, when an individual takes out a loan, she issues a promise to a bank. Once she ‘makes good’ on that promise (i.e. repays the loan), she may ‘destroy’ that loan debt (liability) by eliminating it from her balance sheet. Likewise, the State, once it fulfills its promise to accept its own money (HPM) at State pay-offices, can eliminate an equivalent number of these liabilities from its balance sheet.”
“Thus while bank money (M1) is destroyed when demand deposits are used to pay taxes, the government’s money, HPM, is destroyed as the funds are placed into the Treasury’s account at the Fed. Viewed this way, it can be convincingly argued that the money collected from taxes and bond sales cannot possibly finance the government’s spending. This is because in order to ‘get its hands on’ the proceeds from taxation and bond sales, the government must destroy the money it has collected. Clearly, government spending cannot be financed by money that is destroyed when received in payment to the State!”
Now, Prof. Kelton wrote that in July 1998, and she may have changed her views about this argument since then. But I have seen others reference her paper on this score, and assuming a criticism of this argument has not been given, I will offer one here. But before I do, let me note that her argument seems influenced by Warren Mosler’s book (available for $1.00 at Amazon.com as a download in Kindle format). Mosler tells us:
And what happens if you were to go to your local IRS office to pay your taxes with actual cash? First, you would hand over your pile of currency to the person on duty as payment. Next, he’d count it; give you a receipt and, hopefully, a thank you for helping to pay for social security, interest on the national debt, and the Iraq war. Then, after you, the tax payer, left the room, he’d take that hard-earned cash you just forked over and throw it in a shredder.
Mosler wants to argue that Congress does not need taxes in order to spend. And he seems to want to shock the reader with an astonishing claim. But I’m sure he’s aware that the Treasury keeps an accumulating record of receipts of taxes, fees, fines and outgoes:
That alone suggests that a record is kept of the receipts for taxes given to taxpayers and that these receipts are recorded and added together for monthly and year to date cumulative amounts of tax revenues and other inflows and outflows to and from the Treasury.
Mosler says taxes are needed to get people to use the government’s money and also to regulate inflation. And I think he really means that after checks are sent to the taxpayer’s bank for cashing and cash have been recorded and shredded to save storage costs since they are no longer needed after that, the government still keeps a record of those receipts. But unless you already know that detail, you might conclude what Professor Kelton did, as she attempted in 1998 to remain faithful to MMT theory from one of its leading advocates.
I think she confused the idea of a promised extinguishing of a loan by a promised repayment of it with the idea of a promise by a government to accept its own money in payment of taxes. Loans are usually extinguished when loans are repaid. But acceptance does not imply that a non-loan that is accepted be destroyed. UPS promises to accept your package and deliver it to its destination. The package is not destroyed at the point of acceptance. Acceptance doesn’t contain the meaning of destroying what is accepted. When the government receives money in taxes, it accepts that money with the assumed understanding that the money will be used by the government for something. For that to happen the money must continue to exist in some form after its acceptance. It just does so in a different form than what it came in to the IRS as. Otherwise the government will not know what its deficit is.
The idea that money given the government to pay taxes vanishes as it is accepted and the government then recreates an equal amount with spending authorization and spends that, cannot account for the existence of government surpluses. The surplus is the difference between authorized government spending and tax revenues collected. It is the opposite of a deficit. The money must have continued to exist after being accepted to accumulate in a surplus as the difference between total receipts and spending.
And this idea that the tax money vanishes does not completely accord with the fact that the Treasury keeps this ongoing record of its accumulating receipts. Money is, after all, a quantitative representation in units of account of debt obligations between parties in the economy exchanging goods and services for money. As long as there is a way of keeping count in some record, money continues to exist.
That is the “bean counter” argument, that the records are money and actual money is irrelevant. The same argument allows banks to create money out of thin air by making entries in accounts. The argument also prevents the federal government, which has the tools to manage the money supply (taxing/spending), from adding anything but coins to the money supply, leaving that power to the banking system which does not have tools to directly manage the supply of money in the economy, a fact that has been demonstrated over and over again by cyclic booms and busts.
The issue here is simple. Taxes can only be paid with something that only the government issues, which is called “currency.” In a modern system, currency includes cash and settlement balances which are equivalent. Thus, to pay taxes is necessary to obtain the currency that only the government issues.
Currency can only be obtained from government either 1) from government spending, which is handled by the Treasury and cleared through the central bank. Modern Treasuries direct the central bank to credit accounts rather than the Treasury distributing cash, or 2 by central banks lending to institutions that are members of the payments system run by the central banks. There is no alternative.
When government spends, a non-government bank account is credited, which adds currency in the form of settlement balances to the payments system when the central bank credits the account of the bank of the recipient with settlement balances, whereupon the bank credits the deposit account of the customer. M1 increases, and the monetary base also.
When the customer pays taxes, the draft is cleared through the central bank, whereupon the central banks debits the account of the bank of the customer in the payments system and the bank debits the customer’s deposit account. M1 decreases and also the monetary base.
As a matter of accounting the taxes are credited to the Treasury account at the central bank in form of settlement balances. This is booked as an asset of the Treasury and a liability of the Fed. Net zero. The government has gained nothing. Nor does it need anything since it can issue its own liabilities without limit, the constraint being inflation.
A currency issuing government never needs to obtain funding elsewhere to spend because it spends only its own liabilities, which it alone issues.
“Printing money” as it is used today generally signifies a central bank buying government bonds for settlement balances it creates to do so. This is an asset swap that doesn’t change the amount of non-government aggregate net financial assets, only the composition and term. It increases the monetary base but not the money supply. It’s chiefly a liquidity operation conducted in the payments system, to which only banks have access.
Where a central bank acts as lender of last resort, QE does not affect the money supply in floating rate systems. There is no money multiplier. This was demonstrated in the US, when a huge QE program did not increase bank lending and did not lead to inflation. It did lead to increased equity prices by keeping interest rates low.
It also influences of the yield curve of government securities. When the central bank buys bonds in quantity, it raises the price of bonds and lowers the yield and vice versa in selling. The central bank sets the overnight rate for interbank funds directly in the US (federal funds rate) and delegates it in the UK (Libor) But the central bank can also influence long term rates by purchasing and selling bonds.
Qe has been used in Japan, the US, and Europe to stimulate recovery by increasing bank lending, which is then supposed to stimulate demand in the economy. It hasn’t worked to do this anywhere, so in that regard it’s turned out to be hype. The theory was wrong, as MMT economists said from the get-go.
QE1 and the various facilities the Fed created at the time in the US were useful in relieving the liquidity issues in the banking system that arose when Lehman crashed and the interbank landing froze up because trust broke down. Then the Fed stepped in to provide liquidity and improve confidence. That did work, but the Fed waited too long to do it. As manager of the payments system, the Fed should have acted immediately, but the crisis caught everyone off guard.
The whole QE episode exposed the absurdity of a system where the tools to control the money supply (spending/taxing) are on the fiscal side while the monetary side has the responsibility but not the tools. Recently we see the ECB trying to usurp that spend/taxing power in Greece, putting the tools in the same place as the responsibility. Money gets into the economy when loans are made and it disappears when loans are paid off. With frugality in the public sector, paying off more loans than new ones being made, reduces the money in the economy, reducing demand, shrinking the economy more in a positive feedback spiral of instability. All systems, in order to be stable, must have negative feed back and the federal government is the only entity capable of providing the needed negative feedback. We suffer from a monetary system which is a relic of the “gold is money” era that no longer exists.
What do people think of monetary reformists and similarly aligned economists, seeming inability to engage with and understand the political hurdles to enacting these economic policies?
We seem to be in a situation, where economically, we’ve figured out a great deal of good options for reforming our economies/societies – but in the same community, there seems to be ZERO knowledge of the political hurdles in place in front of us, or how to go about tackling them.
A good example of this, is Yanis Varoufakis’ time as Finance Minister – an excellent economist, with exactly the right ideas for reforming Greece/Europe, but who displayed extraordinary political naivety and proved to be incompetent politically (don’t get me wrong though – economically I still hold him in extremely high regard).
We still seem to be decades away from transforming this academic ideas, into real political movements – what do people think? How should we all proceed?
Sorry mate, you’ve been stuck in the mod queue. All we can really do is educate people on how the systems work.
No worries 🙂 The trouble I find is: Nobody cares about economics – it’s like trying to explain the finer points of quantum mechanics or such to someone, people just can’t grok it unless they’ve been studying it themselves for a long time (which almost nobody is motivated to do – and those that are, often get intellectually captured, into believing neoliberal memes).
The two most insightful and wise quotes I have ever encountered about banking/monetary systems were from Ezra Pound and Henry Ford. Pound opined that when people generally gained an understanding of monetary systems the effect on society would be equivalent to the changes caused by people becoming literate with the ability to read and write. Henry Ford had a similar reaction from a different vantage point with his statement that it was good that people did not understand what banks do because if they did there would be revolution in the streets before sundown!
If people knew the operational details, they would see that the currency is a public utility and that the financial system is dependent on the currency, which sits at the apex of the hierarchy of money. They would also realize that chartered banks are government franchises. As a result they would demand that fiscal and monetary policy, along with the entire financial system, be operated for public purpose, just like any public utility.
I submit that seigniorage is the single most important word in the lexicon of monetary theory. The federal government takes full seigniorage, both positive and negative, on coins but commercial banks draw interest on the seigniorage of currency. That is the problem with our present system. The federal government should be taking full seigniorage on all currency as well.
In turn I submit that seigniorage has not been adequately defined in any literature I have ever read to have any meaning. It also seems to be a word that has some utility as it means different things in different contexts much like the word ‘structural’ when applied to economics & finance. I think they are the weasel words of the industry as there is never any clarity to the intent behind the words.
“Seigniorage” has a very simple and very easy to understand meaning that is really very obvious. It is just the difference between the face value of a monetary unit less the cost to produce it. A $100 bill cost (or did cost) 9.7 cents to print. The seigniorage is therefore $99.903. That is the expensive kind of money. The computer click kind is even cheaper.
The problem with this, is that I don’t think enough people generally will ever gain an understanding of the monetary system – not until after it has already been changed/simplified (with the catch-22, that the former has to happen before the latter) – I have not encountered a single other topic, which triggers such strong brick-wall/stonewalling cognitive-bias, as topics surrounding money creation do – it’s amazing/fascinating and disheartening.
My favourite economics quote – to counter your own 😉 – is from James Kenneth Galbraith:
“The process by which banks create money is so simple that the mind is repelled.”
Extend that sentiment, to the idea of government use of money creation, and I think that’s the best summary of the entire problem that there is.
It just seems politically and socially naive to me, to believe that MMT will get anywhere, just by educating people – a bit like how Yanis was naive in thinking that if he just had good enough arguments and put them forward well enough, that he could bring the other Euro finance ministers on his side.
Tom Hickey quoted Stan on Open Discussion I (continued here):
“What funds will the primary dealers use to pay back the Fed for its loan? And what funds will the banks use to buy them from the dealers?”
Let’s assume that the dealers don’t have the settlement balances in their accounts and their inventories of Treasuries is depleted so the Fed cannot do the usual OMO or POMO. Then the dealers purchase the securities through their overdraft privilege at the central bank at the discount rate.
The Fed then waits about a week and purchases the securities for its own account by creating settlement balances in the dealers accounts to cover it. The loan is the[n] paid off.
Then Treasury had the settlement balances it needs on time, and the Fed has the securities on its books, just as if Treasury had sold them directly to the Fed.
Note that above Fed has paid the Treasury for deficit spending with money it created out of thin air and totally settlement balances.
Stan now think more should be added to take into the account how the banks get the securities:
Fed, next in OMO, presents primary dealer with the security bought initially by the dealer from the Treasury for sale to private banks but bought in the interim by the Fed. Banks pay with (a) bank money created as a huge loan or (b) settlement balances.
Banks get the security from the dealer.
The issue is with (a) and (b). Tom would think that the dealer is an agent of the Fed, thus must work exclusively in settlement balances. Stan thinks the dealer is at the interface between the government and private sector. The dealer is not an agent of the government but a private corporation that acts sometimes as an agent of the Fed and other times as a private agent to banks and individuals. So, the dealer has to keep two books, government money books (settlement balances) and bank money books, (created by bank loans).
The next issue is whether Treasuries must always be bought with settlement balances.
The argument against settlement balances is that for banks they are difficult to acquire in the values needed for deficit spending. Banks must provide some kind of services to earn these from circulation in the economy. So, for the huge amounts of dollars needed, a lot of effort and work is required.
But then the Treasury is going to roll-over the debt to the banks on the securities. It will do this by swapping mature securities held by banks for new securities issued by the Treasury. No money needs be exchanged in these swaps. In fact the Treasury likely would not have the money to buy back the principal of the securities from the banks, and it would have to issue other securities just to get money from other banks to get any such money.
The effect of rolling over the debt is that the principal on the loan is never paid back. (Only interest is paid, financed in a similar method and rolled over also like the loans via securities on the principal).
So, the banks have to wonder why in the world are they buying these securities that will never return the principal, after all the effort used to get it? In fact the method of financing the deficit through the Fed’s buying the security from the dealer, who initially had bought it from the Treasury with an overdraft that the Fed’s buying remunerated with settlement balances made out of thin air.
The argument in favor of settlement balances nevertheless regards the securities as issued by a government agency which must be remunerated with government settlement balances.
It might be possible to work out another way for the private banks to get settlement balances in the quantities needed to buy deficit securities.
Perhaps something analogous to the way the Fed exchanges reserves for cash from the mint can be arranged. (Tom?)
Tom and I have been thrusting and parrying over this at Open Discussion on MMT I, with little participation from others. Perhaps we can get some others involved in this, since there is a lot of holes of undocumented and unverified opinion in the debate, and others may know how to fill those holes with fact.
I have been studying Tom’s account of how the Fed funds the deficit spending indirectly through primary dealers at the Fed’s public auction.
It seems to work.
So, we initially have three parties, Treasury, Dealers and Fed.
Treasury issues securities.
Dealers buy them with overdrafts from the Fed (Fed loans Dealer).
Treasury gives securities to primary dealers.
Dealer then turns around and sells the securities to the Fed, which buys them (as usual) with money it creates out of thin air in the form of settlement balances (Federal Reserve Notes). This cancels the overdraft loan, or the Dealer can use money from the Fed to cancel its overdraft loan from Fed.
This is a point at which actions may terminate. But has Treasury borrowed
from banks in its sale?
Not for the deficit spending.
So is this legal? Or isn’t it?
Are primary dealers creatures of the banks? Or are they private corporations? Or are they agents of the Fed?
Is the dealers’ money for the initial purchase of the securities the Fed’s money instead of private bank money? So, who did the Treasury borrow from for the initial sale of the securities used to fund deficit spending? Are dealers private financial institutions?
On the surface it looks like Fed money all the way to the end of this sequence of moves.
Fed can turn around and now request that the primary dealers put its recently purchased securities up to public auction with the banks.
Dealers don’t buy them, but act as custodial agents in making the sale.
So, as I raised the question above, if the banks have to buy securities from Treasury in settlement balance dollars (FR Notes?), how could they get enough of these through bank services to buy securities for the large deficit spending? And why would they buy them knowing that they would never get their principal back? (Treasury and banks will agree to let Treasury roll over the securities by swapping new securities for the banks’ mature securities, adding in the necessary interest from other Treasury accounts created by additional borrowing from banks for the interest money. These create debts that will also be rolled over just as the principal is. The interest money is debt free as far as the banks are concerned. They can spend it on whatever.
Now we know earlier from inspection that banks have Treasury securities and it is most likely these were bought from the Fed via the dealers at OMO.
Is it possible for the banks to create the money for the purchase of the securities by making a loan to the Treasury with which the banks can purchase the securities (IOU’s from Treasury). So, before sending the dollars to the Treasury, the banks need to convert them to settlement balance dollars.
Is there a way for the banks to get these from the Fed, like they get cash from the Fed, which converts their bank dollars to settlement balance dollars.
Every time a bank creates a loan, and that money (bank money) is taken by a borrower and withdrawn from the bank in the form of cash, a conversion from bank money to settlement balance dollars (in the form of cash) had to be made.
Or can the dealers just switch hats from government to private institutions. Switch to bank-money ledgers and simply sell the objects (securities) to dealers for bank-dollars created by bank loans? Everything is handled on bank-money ledgers.
Then the Treasury rolls over the debt on the sold securities (as above), adding new interest money at the roll-overs (as above). The banks don’t care about the repayment of the principal because the money was just created out of thin air with no effort at all worth speaking. The banks just want the interest.
One other question, Tom, is there official documentation, or first-hand witness accounts (like from primary dealers) which substantiates the scenario you describe–which seems partially plausible, but still suffers from the problem that the Treasury can’t borrow or sell directly to the Fed.
I don’t care whether it is classical MMT or not; I want to know exactly how the system works within the current legal structure. Does it operate like a form of MMT?
To me a much more plausible account is that private banks, like the Fed are sanctioned by the Federal government to create money out of thin air when they loan. They cannot spend money as if it is their own created out of thin air. For example, the banking system in the UK puts banks on an equal footing with the cb when it comes to the ability to create money out of thin air using loans. So, in their system, an MMT account would have to accord this as an independent source of new money. In fact, it is claimed that around 97% of the British pounds are created by banks and not the central bank (cb).
What can decide which is the case would be documentary evidence from Fed, primary dealers, private banks.
Stan, I have explained all this previously from a variety of angles.
To summarize, there are basically four elements in the US system. Other countries systems may be somewhat different but the principles of central banking are essentially the same so that the central banks of different countries mesh at the Bank for International Settlements (BIS)
Divide the system into government and non-government. Divide government into the Treasury and central bank (Fed), Divide non-government into those entities that have access to the payments system directly (banks, foreign central banks) and those that don’t have access to the payments system directly. Divide those that have access to the payments system directly into the primary dealers and the rest.
The Fed and the primary dealers are the principal participants in auctions, the Fed as the agent of the Treasury that sells for the Treasury account and the primary dealer that act as brokers for those who subscribe to the auctions. The primary dealers take the residue of auctions that are undersubscribed for there inventories, which they expand and contract with changing market conditions.
The Fed also expands and contracts its balance sheet using OMO and POMO by purchasing and selling Treasury securities.
Treasury auctions settle in settlement balances (Fed liabilities) in the payments system operated by the Fed. Those that don’t have access to the payments system directly buy Treasuries at auction from those that do.
Customers with deposit accounts at banks hold deposits that are bank liabilities, not settlement balances. Banks promise to convert deposits to currency in the form of cash at the window on demand or settlement balances as needed, that is, for all transaction that don’t involve netting. Most transactions are netted. Banks cannot issue currency so they have to obtain it as cash or settlement balances ultimately from the issuer, the government, through the central bank (Fed).
Banks get cash by exchanging settlement balances at the Fed for it. Banks get settlement balances either from excess reserves in their accounts at the Fed, borrowing from other banks that have excess reserves, or borrowing from the Fed. Most settlement balance enter the payments system by Treasury spending.
The Fed ensures that there are always sufficient settlement balances to clear while maintaining its targeted rate. To do this it expands and contracts the monetary base by buying and selling Treasuries itself. For example, during QE the Fed purchased about a trillion dollars worth of Treasuries just by crediting accounts with settlement balances it creates “out of thin air” by keystroking its spreadsheet.
(Those settlement balances held as bank assets can be exchange at the Fed for dollar bills on demand, just as customers can demand cash at the window. The difference is that the Fed issues the currency (settlement balance and cash) while the bank does not. It can create dollar entries in deposit accounts by keystroking its spreadsheet but it cannot create the currency with which it promises to settle, either at the window in cash or in the payments system in settlement balances.)
There is no problem with this if one follows the accounting all transactions and notes whose books the entries appear on. Where there is overlap is between the books of the central bank and the banks, since that is the interface between government and non-government. When a Treasury is bought by the Fed, the Fed’s security account (asset) is marked up and so its liability account. The settlement balances (Fed liability) are credit to the selling bank’s account at the Fed. The bank reflects the transaction on its books by debiting its securities (asset) account and crediting it settlement balance (asset) account. Thus it is a wash for the bank and the Fed has increased both its assets and liabilities.
If the Treasury securities were sold by the banks for a bank customer then the bank instead marks down the customer’s security balance and marks up the deposit balance, which is a switch of liabilities for the bank and assets for the customer.
Tom, you are not dealing with the key criticisms of my argument:
(1) Your method by which deficit spending money is funded by the Fed sounds plausible, but I’m still a bit doubtful by how close this is to the Fed actually buying the securities directly from the Treasury, which I think is unlawful. If the usual role of the primary dealer is as an intermediary agent who facilitates exchanges, then this looks like the Fed first lending the dealer the money to buy the security from the Treasury and then the Fed buying the security itself from the dealer once he gets it, to give him new money that cancels his debt to the Fed when the security is given up to the Fed. That’s an unusual kind of exchange, but it may be plausible–if only illegal if it is done with the Treasury. To me that amounts to the same thing as a direct sale of the Treasury to the Fed.
It is a procedure that an entity like the Fed, that can create money out of thin air in unlimited quantities and on any occasion, could perform to make it look like it is not directly buying something from someone. That is going to look bad when examined closely.
So, deal with this argument one way or another but don’t shift focus onto a description of the whole government/banking system that ignores this. I agree with most of your description of it. I’m trying to see why it works or is legal.
Next the Fed now starts out with that security it bought from the dealer. Why do we go on to sell that security to the banks, especially if this was done during a recession (as deficit spendings usually occur then)? Why doesn’t the Fed hold on to that security until there is inflation, and then sell it, to counter inflation? Yet, we observe that the amount of US securities now held by the banks is on the order of magnitude of the previous fiscal year’s deficit plus interest on them and previous year’s securities still rolled over. So, for some reason the banks are buying the securities from (?) while the Fed has been buying up the securities from previous years with QE, so we don’t see a proportionately large amount in US securities from previous years now at the banks. (The national debt for deficit spending is getting paid off).
And I have to remind everyone here, if after the deficit spending portion of the national debt has been essentially cleared by the Fed, the national debt still looks huge, it has nothing to do with deficit spending. About 87% of the public national debt is in the hands of investors. Their dollars are essentially tied up in time-deposits (in the form of their securities) held at the Fed. None of that money pays for government spending on various programs and operations. The money is still there to pay off this part of the national debt. A lot of those dollars are a result of years of negative trade balances that have been parked in time-deposit savings for safe holding at the Fed.
Bottom line: the national debt problem is a myth, an illusion because most people, even very smart people don’t know nor understand federal finances.
So, back to Tom and my debate: Tom, do you have documented evidence that what you describe is going on with deficit spending at the dealers and the Fed? Or if you do have it, do you fear for your life if you reveal it? :-).
(2) Why would banks want to buy the securities with settlement balances taken from current circulation, since to acquire them would take considerable effort to provide some service. Name me some services that would raise that amount of money. We already know that the Fed has funded the deficit spending with settlement balances, so the banks are not needed for that. Then to simply use these settlement balance dollars to buy the securities, knowing that the principal will never be repaid, because the debt will be rolled over repeatedly forever, means the banks should think a great deal of lost effort has resulted with little to show for it (only the interest, paid at each roll-over). The Fed has already paid with debt-free dollars for the securities when it bought the securities inititially. It could just as easily swap the mature security it bought for a new one at the Treasury, then sell the new one to fight inflation (should it arise).
Suppose there is a depression or recession going on. Why would banks want to buy the securities then, knowing that the usual procedure was the Treasury rolling over the debt on the securities? The banks would hope for a continuing QE from the Fed to restore their principal on the securities to them. Otherwise there is no full recovery.
Consider further that with the Fed’s having funded in full the deficit spending in your scenario, the withdrawal of the settlement balances in circulation for the banks’ purchase of these securities to an equivalent amount means the money supply has not really increased. And that is not what we want for deficit spending. Deficit spending should increase the money supply accordingly. Maybe QE actually but indirectly increases the money supply when the Fed buys these securities from the banks, since these settlement balances that were withdrawn from circulation can now go back into circulation as a result of QE. That will depend on whether the banks try to do it with loans or just by spending the profits.
Time out to reflect on information often gives me a better understanding of it.
Senexx, I often find my best thinking is done in the shower.
Stan, I am simply following the example of the MMT economists when someone is misconceptualizing operations by stating the operational description.
See Modern Monetary Theory: A Debate
Another reason I don’t answer some questions is that I don’t have spare time to spend trying to figure out what someone are driving at when they are using non-standard terminology. This is what I was told when I a first entered the arena of finance and economics as a newbie. I recall someone telling me that I needed to spend the weekend reading an introductory economics text to get the terminology right since I was obviously confused.
There are several intro books online for free. Randy Wray’s Primer at New Economic Perspectives
and Bill Mitchell and Randy Wray’s draft for their forthcoming macro text.
Both NEP and Bill’s billyblog are treasure chests of intro MMT.
There are other MMT oriented blog where there is discussion in the comments, in particular Mike Norman Economics and Neil Wilson’s 3spoken.
One further thought on Tom’s method by which the Fed funds deficit spending:
Tom’s method, if I understand it properly, has the Treasury issuing securities, the primary dealers buying these at auction with loans from the Fed in the form of overdraft withdrawals from the Fed.
Then the Treasury spends the money from the sale of the securities into the banking system.
A week or so later, the Fed then buys the securities for deficit spending from the dealers with money created out of thin air. This provides the dealers with money to pay back the loan from the Fed to buy the deficit-spending securities.
The Fed now has the securities for deficit spending.
Once the Fed acquires the securities it used to fund deficit spending, instead of next selling these directly to the banks (why they would want to buy them is not clear), suppose the Fed just swaps the securities it bought for new securities from the Treasury. The securities for deficit spending are extinguished when received by the Treasury. The Fed waits until there is inflation to sell these new securities to banks (ordered to buy them to drain reserves into time deposit accounts to limit lending) and investors (to drain money from circulation into time-deposit accounts).
And there is still no need for a platinum coin. The national debt is either paid off by the Fed in QE, rolled over and over by the Treasury, or paid off at maturity in the case of money in time-deposit investor accounts with money from the time-deposit accounts plus interest created out of thin air by the Fed, or paid off by the Fed’s purchase of the deficit-spending securities with debt-free money created out of thin air and the subsequent swap for new securities with the Treasury.
The problem is still the debt ceiling. The tsys on the Fed books still count toward the ceiling. The Fed can’t credit the Treasury account without the issuance of new tsys, which is not permitted after the ceiling is reached. This is what the platinum coin serves to get around.
The platinum coin is that and much more. It is a way to take full seigniorage on a large of quantity of money, blocking commercial banks from drawing interest on that seigniorage. And it is even more. Long term use of the coins to pay for government will retire the national debt, crippling the ability of the Fed to control the money supply in the economy, that responsibility going to the government which has the tools to control the money supply (spending and taxing), tools the Fed has never had anyway.
Senexx, have you thought of anything to add to this discussion?
Nothing that I think you would find useful. As an Australian, the discussion of the US system is overdone. The accounting and finance works by and large as described by MMT and it is all quite simple. The ‘elites’ whoever they may be, the educated, politicians, other economists, etc deliberately construct laws on top of this rather simple system to deliberately obscure how simple it is. All done to enslave the masses because when you pick at the threads and undo it all you see how simple it is. Who has time to do that? People are struggling to make a living and have all sorts of commitments that come first. Some Modern Money Theorists are accused of using counter-intuitive language but it is only counter-intuitive to others that have studied economics, MMT choice of words is much closer to lay English – which is how it gained such a rise in popularity. It is much more accessible than standard Austrian, Chicago and Keynesian economic thought which uses economic jargon and what appear to be outmoded models based on past economic situations, not current economic situations.
Some people accuse MMT of not being a political movement which is why they find it lacking in substance – because it doesn’t have a policy prescription for everything but that’s exactly the point. It is the mechanics of something and now we know the mechanics how can we make the engine more efficient. So overall MMT gets attacked based on the people espousing it are crazy with no evidence of that ever provided and no evidence of what MMT presents ever refuted.
So once we know how the system works – we can do what we want with it according to our own political ideology.
Well, the debt ceiling needs to be attacked directly, not with work-arounds. (1) It is unconstitutional, because Congress cannot pass laws that place conditions and limits on its absolute powers as given in the Constitution.
These are the unconditioned powers to pay debts and to borrow. It will take a Constitutional amendment to place conditions and limits on these powers.
(2) There are three kinds of US securities in terms of how they are ‘managed’, meaning, paid off. None of them involve US taxpayers: (a) deficit spending and interest securities are held by banks, and these are rolled-over perpetually, so that neither their principal nor the interest (also borrowed from banks) is paid off. (b) the Fed with Quantitative Easing may just go into the banks and buy their securities, cancelling the Treasury’s debt to the banks and transferring it to the Fed, which will swap these securities when mature for new securities with new future maturity dates. (The Treasury will extinguish the original deficit spending securities when they come into its possession. These will be held until there is inflation and sold to banks and investors to drain reserves and money in circulation into time deposit accounts at the Fed, out of circulation. (c) US securities bought by investors and held in their name in time-deposit accounts at the Fed, with money to pay off the principal in the time-deposit accounts and interest to be created out of thin air by the Fed. None of the investors’ dollars is spent on government operations or programs. These will be paid back on demand to the holders of the securities when the securities mature. Or these securities may also be rolled over similarly to what is done with deficit spending securities. (d) Unmarketable Intragovernmental series securities, the money for redeeming which the Treasury can raise with marketable securities sold to private banks at public auction. The Fed can subsequently buy these securities with dollars it creates out of thin air, cancelling the debt to the banks. Marketable securities held by the Fed can then be swapped for new securities from the Treasury,etc. etc..
So, the national debt has little to do with government spending (only about 4% is) , and the debt-ceiling is being influenced by investor-held securities not deficit spending securities, so it is an unfair and unreasonable method of constraining Congressional spending.
All of these facts can be checked by Congress.
Tom, why do the banks buy the securities when the Fed has already funded the deficit spending? What’s in it for them? They will lose all those settlement balances in the principal that gets rolled over and over forever by the Treasury. If these are bought with settlement balances drawn from circulation from previous deficit spending, that cancels the effect of the current deficit spending increasing the money supply, since the money withdrawn from circulation will equal the money derived from the Fed and deficit-spent by the Treasury into circulation in the economy. So there will be no change in the money in circulation. It would be better that the banks don’t buy these securities from the Fed with settlement dollars unless those dollars are FR Notes newly given to the banks for the purchase.
When banks purchase securities, they exchange settlement balance
for securities that pay higher interest. It may only be a small amount of interest but it adds up when large sums are involved
So banks do this to increase revenue. When they are exchanging settlement balances they hold that are in excess of the requirement for short term securities like T-bills, there is higher return with zero risk. It’s therefore a no-brainer.
Corporations also hold “cash” mostly in the form of short term government securities rather than deposit accounts for the same reason.
No one ties up principal in holding tsys since they can be exchanged for settlement balance by selling them in the market, and the tsy market is highly liquid.
Tom, your answers here sound plausible. I failed to recognize that under your account the banks don’t buy the securities to finance the deficit. (I refused to believe that the Fed was effectively buying securities from the Treasury but doing it in an indirect manner. Most accounts say that the Treasury has to borrow from the banks to do deficit spending. Even Frank N Newman teaches this in his books when he discusses how Treasury rolls over the debt). Now that I understand it and feel comfortable with it, I can accept it).
The Fed finances the deficit but indirectly through a special ruse using an intermediary (the primary dealers). The banks, then, are just buying the securities as investors, and their dollars spent for the securities (principal) will continue to exist in time deposits at the Fed, which will be returned to the banks on demand at maturity. So, that makes investing in these securities for the interest a reasonable financial act, since you get your principal back also.
Still, for the interval in which the banks’ investment in the securities remains in the time-deposits at the Fed, they have been taken out of circulation, which can be deflationary, unless we can suppose they were simply transfers of funds already out of circulation, from bank savings to savings at the Fed. (Remember the Fed sells securities to counter inflation, so Fed selling these securities is deflationary). So, it may be that the Fed will hold on for awhile until inflation develops before selling these securities to the banks.
Still, we must recognize that banks’ buying securities equal in value to those issued by the Treasury for deficit spending, but bought by the Fed, will cancel the Fed’s additive effect on the money supply from its financing deficit spending, until these securities mature and the banks get their money back to spend.
I still have to wonder why it is that the banks still have securities, after the Fed has pillaged them for securities with QE, equal approximately to the previous fiscal year’s deficit plus securities issued for interest borrowing by Treasury. Maybe it is just a one-time coincidence in the data I saw (2010 and 2014 data on who owns the national debt).
Now, can you get someone at the Fed to confirm your account of their operations, that they indeed are the financier of the deficit spending but indirect in manner of doing it. It has been very frustrating for me to try to learn exactly what the Fed is doing to manage deficit spending and the national debt.
Anyway, with the Fed more or less directly canceling the debt for deficit spending by buying the securities (from primary dealers) before they are sold to the banks, that eliminates a major reason to reject the debt-ceiling and its use of the amount of the national debt as irrational and irrelevant limits on deficit spending.
Perhaps the Fed could just swap the deficit-spending securities for new securities before selling them to the banks. Then the debt obligation explicitly stated in the new securities no longer is for the deficit. The deficit-spending securities are themselves extinguished when they get back to the Treasury. But the new securities are not for the deficit, which has been cleared by the Fed.
I would like very much to hear comments from all of you here on the proposal described at http://www.realmoneyecon.org. It is supported by an array of experts and is based on the “Chicago Plan” unveiled originally in 1935.
Stan, I don’t have time at the moment to go through all the points in your post, but the salient point is a misunderstanding.
” (Remember the Fed sells securities to counter inflation, so Fed selling these securities is deflationary). ”
The Treasury taxes to run a balanced budget or surplus. When the Treasury’s balance is at the point of deficit, then it directs the Fed to auction some tsys that the Treasury issues for the purpose. The Fed auctions the securities into the market using the primary dealers as brokers, exchanging tsys for settlement balances. This show up on the Fed’s spreadsheet as a debit to bank’s deposit accounts at the Fed and a credit to their tsys accounts, which can be conceptualized as transferring a deposit to a CD in the banking system. The auction is paid for by transferring settlement balances from the private sector (non-government purchasers of tsys) to the public sector (credit to Treasury account).
The Treasury then directs the Fed to credit banks accounts in the name of the recipients of the Treasury spending and transfers. So in aggregate, the amount of settlement balances remains the same in aggregate when all accounts are settled, and non-government holds newly issued tsys in the amount of the auction, which corresponds to the deficit that was funded by the auction. So non-government then has the same amount of settlement balances as before, and also the newly issued securities, which in an increase in non-government net financial assets in aggregate.
The Fed doesn’t need to purchase tsys indirectly to finance the deficit, although it could do that if the auction was undersubscribed and the dealers did not wish to expand their inventories long term. The Fed can let the dealers take the inventory by lending agains the issue and quickly buying it back for its inventory. In this way, the Fed can fund the deficit, which is in effect the government funding its deficit with securities instead of directly issuing currency into non-government, e.g., as greenbacks.
Most people think that securities issuance “sterilizes” the amount of spending but actually it doesn’t. Savings bonds that are held by the purchaser until maturity would, but tsys are the most highly liquid financial vehicle other than cash. T-bills are essentially cash equivalents. Anyone can sell any amount of tsys into the market and the market will always clear because the Fed will ensure that it does by stepping in with unlimited purchasing power since it issues the settlement balances that along purchase tsys.
The way the Fed targets inflation is by raising the policy rate, and its does this when non paying IOR, in which case it sets the rate directly, by adjusting the monetary base so that fewer excess settlement balances are available, which drives up bidding for them. The higher borrowing rate is supposed to make borrowing from banks more expensive since they pass costs on to borrowers. The higher rates decrease firm investment and household consumption and the economy slows down, reducing the pressure on wage increases, which is believed to be the primary cause of inflation.
Stan: ” (Remember the Fed sells securities to counter inflation, so Fed selling these securities is deflationary). ”
The Treasury taxes to run a balanced budget or surplus. When the Treasury’s balance is at the point of deficit, then it directs the Fed to auction some tsys that the Treasury issues for the purpose. The Fed auctions the securities into the market using the primary dealers as brokers, exchanging tsys for settlement balances.
Tom, are you now discussing the sale of securities the Fed has bought through the primary dealers to banks? I am a bit confused because you begin with the way deficit spending occurs and then seem to discuss how the dealers are selling the securities to the banks as if they are the initial buyers of them.
I think your original description of how the Fed finances the deficit by the Fed providing overdraft funds for the primary dealers to buy the Tsy’s securities and then a week later buys those same securities from the dealers with money it creates out of nothing and cancels the overdraft debt when the dealers apply that money to the debt is the really new idea for me now and I want to work with it. So, please do not mean to abandon this account after first presenting it.
So, I will try to interpret your posting to concern whether or not the subsequent second sale of the securities, this time to the banks, is deflationary or not.
You have to understand that I interpret things in terms of my framework involving flows in and out of a sector (circulation), and inflation and deflation for me is understood in terms of whether there is a sufficient amount of money circulating in circulation or too much or too little, to sustain full production and employment at stable prices and wages. To me that is the whole system we need to be thinking about.
I am interpreting your terminology of ‘private sector’ and “public sector’
(a little pedantic, I might say, but perhaps the way it is often expressed when everyone knows to which and whom it refers) to mainly refer here to the banks and, if you specify it, investors. But as I now see it the banks are just investors here and not financiers of the deficit.
The Fed, under your account is the financier of the deficit. It pays for the deficit in its entirety. The Tsy does not have to pay back the Fed for the money it gets: the deficit money is now debt-free; it is not a loan from the Fed. However, if and when the Fed swaps these securities for deficit spending for new securities from Tsy, that extinguishes finally the government’s debt on the securities to pay the bearer on demand at maturity, which Tsy does with the new security of equivalent value.
This show up on the Fed’s spreadsheet as a debit to bank’s deposit accounts at the Fed and a credit to their tsys accounts, which can be conceptualized as transferring a deposit to a CD in the banking system. The auction is paid for by transferring settlement balances from the private sector (non-government purchasers of tsys) to the public sector (credit to Treasury account).
The Fed doesn’t need to purchase tsys indirectly to finance the deficit, although it could do that if the auction was undersubscribed and the dealers did not wish to expand their inventories long term. The Fed can let the dealers take the inventory by lending agains the issue and quickly buying it back for its inventory. In this way, the Fed can fund the deficit, which is in effect the government funding its deficit with securities instead of directly issuing currency into non-government, e.g., as greenbacks.
What did you mean above by ‘lending against the issue’. Don’t the dealers pay the Treasury then for the securities? Or is the Fed just using the securities as collateral for the loan? Who owns the securities? The dealers?
Can dealers treat them as their own without paying Treasury for them, that is treat the securities as their own as collateral for the loan? I doubt it. Dealers pay the Treasury for them (buy them) with dollars lent them from the Fed. Then they can sell them to the Fed, and the Fed gives the dealers the money for them, and they in turn pay back the loan with the dollars they got for the securities.
Bankers do some marvelous but magicky things.
Still, what you are describing is that the Fed is lending the dealers the money with which to buy the securities from the Treasury; they pay the Treasury for the securities and then the Fed ends up by buying them with money it again creates out of thin air, but this time for a purchase and not a loan. This provides the dealers the money they need to pay back the temporary loan from the Fed for their purchase of the securities from the Treasury. And as you say, this “…is in effect the government funding its deficit with securities instead of directly issuing currency into non-government as greenbacks.
Well, the Treasury is not issuing currency into the deficit spending, which would be ‘greenbacks’. It is issuing securities. But in return for them it gets FR Notes (or reserve balances) which get issued into the economy/banking system as deficit spending.
Securities are exchangeable for dollars, but themselves cannot be spent for numerous things and varying amounts. Is there a term for this? If I have a $5B security, I cannot use it to buy a hamburger at Wendy’s and groceries at Publix and a book from Amazon, and expect change. I need dollars and change.
There are numerous uses for T securities, most as investments.
My thinking that for the securities initially bought by primary dealers, then bought by the Fed, and then given to the dealers to sell to the ‘private sector’, we have to consider that money leaving ‘circulation’ to savings, as time-deposits somewhere, as imports, is deflationary unless circulation is resupplied with money (dollars) from some other external source: exports, investment, government spending to compensate. The securities in themselves are not deflationary. The money behind them, if taken from circulation for some extended time, produces deflation.
Hence, keep in mind that the securities from the private sector (I will try to use your terminology) will have been circulating for some time since they were created by the Fed and then spent much earlier into circulation. Then when they are used to buy the Treasury securities, and the money for them is then deposited in time-deposit accounts (since they are not used to fund government operations), they are withdrawn from circulation, reducing the amount of the money supply in circulation. This could be deflationary.
However, the money for deficit spending could be inflationary. If the amount of money taken from circulation by banks is used to buy the securities, they will be stored by the Treasury in time-deposit accounts at the Fed. These are not in circulation, as any savings account is not in circulation, until money is readily withdrawable from them.
Tom, the effects of the Fed’s raising interest rates is to reduce borrowing and lending, so that the amount of money in circulation does not grow. But you have to withdraw money from circulation itself to counter inflation. And when the Fed sells securities to investors and banks (who here are also investors), that withdraws money from circulation. That is because the banks’ buying the securities has their investor dollars stored in time-deposit accounts, out of circulation. That is how you deflate an inflation. Do you agree?
My recommendation would be for the Fed to sell its securities when there is inflation or potential inflation. Buy securities when there is deflation.
Additionally, Congress should only deficit spend when there is a recesssion, or a need to grow the economy (population growth, war, or new resources for increasing production without inflation).
The Fed provide the currency that funds the deficit but that is not the usual way it happens. The Fed has the ability, but there is seldom the need to use it.
The Treasury funds the deficit itself indirectly in this way. The Treasury issues tsys which are sold at auction. The payment for the tsys drains existing settlement balances from the payments system, and if there are insufficient settlement balances, then the Fed adds them through OMO. The Treasury adds the same amount of settlement balances used to purchase the tsys through its deficit spending. As far as settlement balances go, it’s a wash in aggregate.
According to one theory, the auction could fail to clear because the private sector doesn’t desire to hold tsys. In this very highly unlikely event, then the Fed can fund the auction itself as I have shown operationally.
But US tsys are always in demand so the Fed having to resort to funding the deficit by the backdoor is unlikely. But there are countries in which this could happen, as it Russia at the time of the default. The Central Bank of Russia could have handled the situation as I indicated but didn’t and Russia defaulted. This was a voluntary default, since they could have avoided it using the power of the central bank to create currency.
Tom said: The Treasury funds the deficit itself indirectly in this way[:] The Treasury issues tsys which are sold at auction.
Tom, you are glossing over too much here. Sold to whom?
The payment for the tsys drains existing settlement balances from the payments system, and if there are insufficient settlement balances, then the Fed adds them through OMO.
How does the Fed add these with OMO? Buy, sell securities? Whose? To whom?
The Treasury adds the same amount of settlement balances used to purchase the tsys through its deficit spending. As far as settlement balances go, it’s a wash in aggregate.
In other words, the Treasury has to deficit spend with money obtained from elsewhere as settlement balances, no? Are you assuming (again too much glossing over here) that the Fed started the auction by lending overdraft money to primary dealers, who then bought the deficit-spending securities from the Treasury? The Treasury takes the overdraft money from the dealers and in return gives them the treasuries. Next, the Treasury deficit-spends this money into the economy (ultimately into the banking system).
This is money never put into circulation before.
It will be debt-free depending on who and how the securities now held by the primary dealers is handled, i.e. who buys them next?
If the Fed buys them with new money created out of thin air, this cancels the loan to the primary dealers, and this new money is extinguished (but the original money from the overdraft still exists and is now in circulation). This makes the deficit-spending money debt-free. However, there is, independent now of this deficit-spending money, the debt obligation of the government to pay the bearer of the securities issued for deficit-spending face value in dollars of the securities on demand at maturity. The Fed can swap these mature securities for new securities from the Treasury. When the Treasury gets them, it can extinguish them. Then the deficit-spending money is completely debt-free.
But even after the Fed funds the deficit spending with settlement balance dollars from its overdraft loan to the primary dealers, a different scenario can develop wherein private banks buy the securities from the primary dealers without the Fed intervening further, for the time being.
The banks’ settlement balance dollars are given to the primary dealers, who in turn use them to redeem the loan for the overdraft by giving these dollars to the Fed. That cancels this debt and these dollars are extinguished as a result. So, there are no bank held settlement balance dollars now to deposit in time-deposit accounts at the Fed. This forces the Treasury to resort to perpetually rolling over the debt to the banks, as a way of managing this debt. The banks realize they will not get their principal back, even as they get back interest at each roll-over. Maybe they will not want to buy these securities if they know that the primary dealers originally bought them with overdraft money from the Fed. They may prefer dealing directly with the Treasury, via the primary dealers as simply neutral agents of the auction and not just owners of the securities.
However, this may have the effect of not increasing the money supply in circulation.
Why? When the banks buy the securities directly from the Treasury (meaning there is no intervening owner), their settlement balance dollars are taken out of circulation and then inserted back again with the deficit spending. So, there is no net change in dollars in circulation. The Treasury will have to roll-over these securities, perpetually, with swaps of new securities for old (no money needs changing hands) plus interest (also financed by issuing and selling securities and managed in the same way).
But the banks similarly will know they will not get their principal back, and the interest may not be sufficient. Only if the banks could create the money out of thin air to buy the securities would it be a sound investment, since nothing previously earned would be lost, and the interest would be debt-free. (It is debt free because the principal will never be paid back as long as the debt is rolled over).
If there were a mechanism by which bank created dollars (through creating loans) could be converted into settlement dollars at par, perhaps only for the purpose of buying Treasury securities, then it would seem advantageous for banks to buy the securities with ultimately loan money for lending to the Treasury.
We have already considered another scenario: Suppose the Fed buys the series of securities for deficit spending from the primary dealers, after they have bought them from the Treasury with overdraft money from the Fed.
The dealers in turn redeem their debt with the same money they get from the Fed for the securities. And then, the best move to make next would be the Fed swapping these securities for new ones from the Treasury. The Treasury then extinguishes the deficit-spending securities. The new securities now can be held until there is inflation, at which time the Fed may sell them to investors and banks to drain money from circulation, reducing the excess money (if there is already full production and full employment), until the amount of money in circulation is just sufficient to maintain full production and full employment at stable prices and wages.
If the Fed turns around after buying these securities from the primary dealers and offers these same securities up for sale and the banks buy them, this has no effect on deficit-spending, since the Fed has totally financed it indirectly. The banks’ money is this time deposited in time deposits like other investors dollars, and held there until the securities mature, and depending on whether inflation is continuing or not, they may be rolled over, or the money in the time-deposit accounts returned to the banks with interest (created by the Fed out of thin air).
It would seem that the best procedure for financing deficit spending would be for the Fed to lend money to primary dealers and then buy the securities they get for them from the dealers and swap these for new securities with Treasury. Then any sale of securities to banks can be treated as ordinary investor sales, with the money for the securities being stored in time deposit accounts to be returned on demand at maturity or rolled over by mutual agreement. This procedure results in the deficit-spending money being not just new additions to the money supply in circulation, but debt-free. Then whether deficit spending’s addition to the money supply in circulation in this case will be diminished, will depend on sales of securities to banks and investors.
“In other words, the Treasury has to deficit spend with money obtained from elsewhere as settlement balances, no? ”
When the Treasury spends it directs the Fed to credit accounts held at the Fed with settlement balances that are only issued by the Fed as its liabilties, just as the only Treasury can issue tsys. The government (Fed and Treasury) only do business with the nongovernment using the government’s own liabilities, that it alone issues. This is the government’s monopoly over its currency. It alone issues it and only conducts its financial affairs using it.
The TT&L accounts the Treasury keeps at the banks simply leaves settlement balances in the monetary base temporarily until they are transferred to the Treasury General Account, at which time they no longer count toward the monetary base. The purpose of the TT&L accounts is for the Fed to manage the policy rate by adjusting the size of the monetary base to affect the overnight rate in the interbank market when the Fed is not paying IOR.
Non-government can only do business with the government — pay taxes, buy tsys, sell to government agencies and get paid by the Treasury using Fed as settlement agent, etc — using the government’s liabilities.Thus nongovernment must either get already existing settlement balances in the payments system from prior Treasury spending, OMO, POMO, or Fed lending. or else either sell tsys in the market to obtain settlement balances, or borrow settlement balances from the Fed.
The uses OMO and POMO to ensure that sufficient settlement balances are available for settlement, including bank overdrafts using the discount window at the penalty rate.
Issues never arise between government and nongovernment because the Fed is the lender of last resort and always can ensure that the payments system clears.
Where issues may arise is within government between the Treasury and Fed since the Fed cannot lend directly to Treasury, that is, the Treasury doesn’t have overdraft privileges at the Fed as other members of the payments system do, and the Fed cannot purchase securities that the Treasury issues directly from the Treasury.
So both Treasury and Fed have to act indirectly through the bond markets. The Fed acts as the Treasury’s agent in conducting auction, and buys and sells for its own account. The Fed uses the agency of the primary dealers that function as the interface between government through the Fed and the private sector as broker/dealers, although the Fed can also conduct business with others in nongovernment, too.
Because the government is the sole provider of its currency as currency issuer, it is never forced to obtain currency as a currency user even though it acts as user in some transactions by obtaining settlement balances and tsys in the nongovernment markets. It is not dependent on this, as Chairman Greenspan testified to Congress, because it is ultimately the currency issuer that can always fund itself directly when needed.
The only issue under the current arrangement is the debt ceiling, where the Treasury not permitted to issue tsys above the limit. This means that the US Government cannot conduct business at that point once the existing account are drawn down. If the US could not pay interest on its debt, then it would be in default of the debt. This is purely a voluntary act by Congress in failing to raise the debt limit. The platinum proof coin is a possible work around, and the president can direct the Treasury secretary to direct the Fed to directly credit government accounts, a power the president already has under emergency power. The president could justify this at the Supreme Court based on the 14th Amendment if Congress challenged his power to do so.
In non-government, transactions are conducted through banks as agents using deposit accounts (checks and electronic transfer) and cash. Deposit accounts are bank liabilities created by banks. To settle in the payments system banks have to use settlement balances, which they can only get ultimately from government spending, the Fed lending against or purchasing tsys, borrowing from other banks in the interbank market, getting transfers of settlement balance from other banks through settlement, and borrowing at the discount window in the overdraft privilege that banks have.
All cash in circulation some from the Fed by the banks exchanging settlement balances obtained as above for vault cash and then passing it on to non-banks. Bank deposits are exchangeable on demand for cash.
So bank promise to exchange their own liabilities, which they alone issue, for government liabilities, which government alone issues. Solvent banks can always get the settlement balances necessary to meet obligations as long as they are members of the Federal Reserve System.
“But the banks similarly will know they will not get their principal back, and the interest may not be sufficient. Only if the banks could create the money out of thin air to buy the securities would it be a sound investment, since nothing previously earned would be lost, and the interest would be debt-free.”
When dealing with IOUs (even cash is a government IOU as a tax credit), there is always the possibility of default, even in the case of governments. It is not avoided by adding bank liabilities since banks are more subject to default than governments.
In addition, inflation is often traceable to excessive bank creation of deposit balances. Many people think that banks already have the privilege of increasing M1 too much.
If bankers could create settlement balances in the payments system, then bankers would be the government. They already are through the revolving door but giving them control directly would be favoring capital over democracy. Now both Fed and Treasury are creatures of Congress, and banks are chartered by laws passed by Congress and regulated by agencies created and supervised by Congress.
As Warren Mosler has pointed out, the simple why to go to “debt-free” money is by setting the policy rate permanently to zero, issuing only T-bills that are essentially cash-substitutes and abolishing the reserve requirement as Canada has done, for example. This is simple to do in the present system, although the financial world would scream at the loss of interest on longer term debt.
The Fed normally doesn’t purchase anything from the private sector other than tsys. When it loans, it generally loans against tsys as collateral. This is a requirement of normal Fed operations.
When banks have insufficient required settlement balances (required reserves) at the end of a period, then the Fed loans the bank the amount it is short at the penalty rate overnight. The bank has to pony up the next day or continue paying the penalty rate. However, just having to access the discount window at the penalty rate shows bad asset & liability management (ALM), so it not only costs the bank higher interest but also carries a cost in reputation.
The Fed has emergency powers, too. In QE the Fed purchased MBS (mortgaged backed securities) from the banks, that is, took the banks’ dodgy assets on its books so that the banks could pass the stress test and not need to raise capital or be put into resolution.
In OMO (open market operations) the Fed adds settlement balances by loaning against tsys as collateral. In POMO (permanent open market operations) the Fed purchases tsys and credits the sellers with settlement balances.
The important takeaway is that the Fed never needs to get settlement balances since it creates them on its own spreadsheet, where they alone exist. Settlement balances are exchangeable at par with cash by banks that are members of the Federal Reserve System.
The Treasury never needs to get tsys anywhere. It creates them as its own liabilities.
The Treasury cannot spend its own liabilities. It can only spend using settlement balances that are liabilities of the Fed. It cannot exchange tsys for settlement balances directly with the Fed, so the dealers are used as a middleman and the dealers act as brokers in distributing the tsys to the private sector. I have explained how the Fed can buy auctions that don’t clear from the dealers in order to ensure that all auctions clear, but this is rarely an issue. Ordinarily auctions are either fully subscribed or the dealers expand their own inventories to clear undersubscribed auctions.
Tom, could you expand a bit on this? In OMO with whom and when does the Fed add settlement balances by loaning against tsys as collateral. Is this something like the way the Fed lends
the settlement balances to the primary dealers using the securities they got as collateral. Then in POMO the Fed purchases the tsys from the dealers and credits them with settlement balances, which they in turn use to pay back the loan.
Does anything else like this occur, say, between banks and Fed? For example, the Fed may loan the banks money to buy deficit spending securities (the collateral); they buy them, and the banks get the securities. Then the Fed buys the securities from the banks (with money made out of thin air), and the banks turn around and give that money back to the Fed in repayment of the loan. That money is also extinguished. The banks also give the Fed the securities, which it later will swap with the Treasury for new securities, which it will sell to banks and investors when inflation arises.
Would the primary dealers canvas the banks to see if any of them would like to buy securities with a Fed loan? Then take it from there.
Is this what they call a ‘repro’? Now that I see how it works, it is very enlightening with how the Fed could help banks buy securities and the deficit spending securities be new settlement balance money and eventually debt-free money. But banks don’t get the long term interest unless the banks hold on to the securities for a number of roll-over cycles.
That opens my eyes now to a whole host of applications of this, which suggests that buying securities for deficit spending can be done in numerous ways.
So, thanks the the ideas. I don’t buy individual stocks, so I hadn’t absorbed that concept of the ‘repro’. But I suppose there is a lot of that over and beyond the Fed, the primary dealers and the banks.
Wikipedia/Repurchase Agreement: “A repurchase agreement, also known as a repo, currency repo, RP, or sale and repurchase agreement, is the sale of securities together with an agreement for the seller to buy back the securities at a later date. The repurchase price should be greater than the original sale price, the difference effectively representing interest, sometimes called the repo rate. The party that originally buys the securities effectively acts as a lender. The original seller is effectively acting as a borrower, using their security as collateral for a secured cash loan at a fixed rate of interest.”
This is the way that banks borrow settlement balances from the Fed at the Federal Funds Rate (FFR), which is the interest rate that the Fed sets as the “own rate” of the currency as currency monopolist, I.e., sole issuer.
The Fed also loans by overnight overdraft in the case that a bank comes up short on its required reserve balance in its settlement balance account at the Fed. This is called borrowing at the discount window. Then the Fed charges the penalty rate, which is generally higher than the FFR.
All member banks have borrowing privileges at the Fed under the conditions set by the Fed.
The Fed also buys government securities in the market by expanding its balance sheet, that is by increasing its assets. This is called POMO, or permanent open market operations.
The Fed generally conducts POMO through the primary dealers at the Federal Reserve Bank of New York.
I’m still confused a bit by the way you do not amplify on certain terms.
What I’d like to know are the detailed conditions. Can the banks borrow from the Fed settlement balances for purchases of securities for deficit spending with an open ended maturity date on their loan from the Fed?
And can they do this without penalty, since they are doing a good deed in buying the securities from the Treasury for deficit spending. After all, when they buy the securities (using settlement balances borrowed from the Fed), the Treasury will take their settlement balance dollars and spend them ‘into the economy’, and the dollars there are no longer directly returnable to the banks to help them resolve the loan from the Fed. Furthermore, the Treasury doesn’t have enough tax revenue to pay back the banks the value of the securities to them (which would render to them settlement balances to pay back the loan from the Fed), so, it and the banks quickly see a common need to put the debt on the securities on ice, so to speak, by agreeing to perpetually roll over the debt with swaps of new securities from the Treasury for mature securities from the banks, plus interest at each swap. The banks are in this situation because they want the interest. The Treasury is in the swapping game because it doesn’t have the settlement balances to buy back the securities. But it needed the principal to cover the deficit. So, it has to roll them over. And it can’t borrow the settlement balances from the Fed.
If the Fed only lent overnight, or for 30 days, 90 days, 120 days, whatever, that might not work for banks buying deficit-spending securities because those securities will be in perpetual roll-over mode. So, will the Fed give banks a break in not charging penalty interest and not putting a date at which the debt has to be resolved, to allow them to earn a bit of interest.
The Fed can always at any time demand surrender of the securities to it to resolve the loan.
“Can the banks borrow from the Fed settlement balances for purchases of securities for deficit spending with an open ended maturity date on their loan from the Fed?”
Loans come with a term. But this is irrelevant since the Fed can roll loans over. After the financial crisis the Fed rolled over loans to the tune of some 30 plus trillion over the course of the crisis.
But the Fed generally makes sufficient settlement balances available in the payments system for banks to borrow from each other, which is the preferred way. The amount of settlement balances is never an issue since the Fed issues them and exchanges them for existing tsys to expand the monetary base as liquidity is needed. The Fed ensures that Treasury auctions clear, just as it ensures that all bank drafts clear. It just uses different tools.
“And can they do this without penalty, since they are doing a good deed in buying the securities from the Treasury for deficit spending. ”
The penalty rate only applies to banks’ overdrafts at the Fed. There should be no need to use overdrafts if the bank manages its asset and liabilities carefully. The overdraft option is there to ensure that the payments system always clears. The Fed supplies the required reserves when a bank is under the requirement at the end of a period.
I accept what you say about the Fed being the origin of the settlement reserve dollars, and that they exist in the federal-level banking system in a finite quantity. It is a neat idea, that they can be used over and over again to shunt private bank dollars into the federal payments system and then to the Tsy, from where they can go to reserve accounts at banks to be exchanged again into bank dollars and placed in accounts of recipients of govt money.
“My recommendation would be for the Fed to sell its securities when there is inflation or potential inflation. Buy securities when there is deflation.
Additionally, Congress should only deficit spend when there is a recesssion, or a need to grow the economy (population growth, war, or new resources for increasing production without inflation).”
Now you are moving beyond monetary operations to economic theory, over which there is considerable controversy. These are issues that the MMT economists have dealt with.
“My recommendation would be for the Fed to sell its securities when there is inflation or potential inflation. Buy securities when there is deflation.”
It is pretty well established empirically that the size of the monetary base is irrelevant to price level. Friedman’s monetarism is essentially discredited, and subsequent versions of monetarism are not in accord with actual operations and their effects. Adjusting the size of the monetary base through OMO functions to set the policy rate, and it is the policy rate that is thought by monetarists to affect price level. Data shows that there are problems with this position empirically, too, which would be expected if operations are misunderstood.
“Additionally, Congress should only deficit spend when there is a recesssion, or a need to grow the economy (population growth, war, or new resources for increasing production without inflation).”
The basic issue of macroeconomics is balancing the trifecta of growth, employment and price level. The ideal is an economic policy that provides for optimal growth at full employment, meaning everyone willing and able to work has a job offer, along with stable prices. As a macro theory MMT provides a solution to the trifecta that most economists think is not resolvable, that is, it not possible to have all three simultaneously. This involves chiefly the sectoral approach to stock-flow consistent (SFC) macro modeling using the sectoral balance approach and fiscal policy based on functional finance, along with a job guarantee to mop up residual unemployment.
This is what macroeconomics with respect to economic policy (political economy) is about. Needless to say, it is controversial, and MMT is one of many positions from laissez-faire at the extreme right and socialism on the extreme left, with some version of a mixed economy in the middle. MMT is a version of a mixed economy that is considered on the left since it emphasizes addressing unemployment. However, its monetary description is neutral, being a description of operations rather than a theory involving economic causation. The whole of MMT is the operational description and the macro theory along with policy recommendations.
Most financial issues, such as asset bubbles and price instability result from bank lending. The MMT solution is to regulate bank lending. Bank lending is based on liquifying collateral, so the place to regulate banking is credit conditions to prevent imprudent lending practices before they start. MMT doesn’t see monetary policy as the best way to address price level since it is a blunt instrument that uses unemployment as a tool to address inflation.
That is, when the Fed detects the potential for wage pressure rising, it raises the interest rate to contract the rate of growth of the economy. This results in a permanent buffer stock of unemployed that expands and contracts with the business cycle but true full employment defined as a job offer for anyone willing and able to work is never achieved and is in fact ruled out as a matter of policy. MMT replaces the buffer stock of unemployed with a buffer stock of employed that expands and contracts cyclically.
First, I want to thank you, Tom, for your engaging me in this discussion. I have learned a lot.
Second, I had a thought last night involving a slightly different scenario.Originallly your scenario was the one where the Fed loans money (created out of thin air) to the primary dealers to cover their payment of Treasury for the deficit-spending securities, then a week later the Fed buys the securities from the dealers with money it again creates out of thin air. The dealers then turn around and pay the Fed for the overdraft loan. The money from the Fed in the second round is extinguished as the dealers pay off the overdraft loan with this money from the Fed.
Suppose, instead, after the primary dealers buy the securities with the overdraft loan from the Fed, the dealers then sell them to the banks to get the money to pay back the overdraft loan. (Can they do this? Is it theirs to sell only after they have paid off the overdraft loan from the Fed and have clear title?) But suppose they have. I see subsequent problems.
If instead of the banks’ settlement balance dollars going into a time-deposit account at the Fed, they are sent by the dealers to the Fed to repay the loan
That extinguished these dollars.
So, the banks have the securities but the Treasury does not have the money to pay the debt obligation on the securities to whomever holds them.
That implies that the Treasury is going to go into a perpetual swapping roll-over scenario with the banks, paying them not principal but only interest that they similarly finance by borrowing from banks with securities and doing roll overs with them.
This is not the best situation for the banks, since they would rather have had their dollars just put into time deposit accounts, knowing it will be returned to them at maturity on the securities along with interest. Now they would only get the interest, and that might not be enough to compensate for the loss of the principal that will never be paid back.
What the banks will have to hope for is another QE, where the Fed intervenes between banks and Treasury to buy directly the securities possessed by the banks. That returns the equivalent of the principal plus interest and cancels the loan of Treasury to banks in these securities. The Fed can take the securities it bought from the banks and swap them (usually after they mature) for new securities with Treasury. This returns the securities to the Treasury which can extinguish the debt obligation to anyone who holds them to pay them face value of the security.
In the meantime the settlement balance dollars used by the banks to buy the securities have gone out of existence and out of circulation. So, this might be potentially deflationary.
Everyone gets paid principal and interest on time. The principal is returned when the tsys matures and is redeemed by the Treasury. The rollover is just the purchase of newly issued tsys that replaces those that are redeemed.
The same party doesn’t have to purchase the new issue for a rollover. This happens in aggregate. The rollover here just means that the Treasury issues new securities that get purchased by the settlement balances added in aggregate from the redemption. There is no increase or decrease of non-government net financial assets in aggregate. In other words it’s a wash. It’s going on all the time and no one ever notices unless one is examining the daily Treasury statement.
The Treasury can go on rolling over its debt forever without ever paying down the debt, let alone paying it off.
Reducing the outstanding debt means decreasing the net financial saving in aggregate of non-government. It removes wealth from non-government and the government gets nothing substantial that it didn’t already possess as the currency issuer. When that has happened in the past it has been followed by recession or depression.
The currency issuer doesn’t have to get currency from anywhere but itself, since it is the sole issuer.
I would assume the process you describe is effected by a rather simple computer program using a data base of tsys, due date, interest rate and bank account of the owner. On the due date the owners account is increased by the correct amount and that amount is added to a tsys “to be sold” data base. Can you affirm the process is all automated in this fashion? I can’t imagine it being done manually in this day and age.
All government and firm bookkeeping has been done digitally for some time. It would be a huge burden to do it manually, involving high costs.
The push is now on for digital money to replace cash. Greater efficiency results in reduced transaction cost. And digitization puts all transactions on someone’s books. Then there is no place to hide.
Exactly what the spreadsheet looks like is, I believe, proprietary. Information about details is distributed on a need to know basis.
Central banks generally don’t say a lot about how they conduct their business day to day. What is known is based on information provided by them and the accounting records they publish periodically.
For example, it is known that the Fed and Treasury are in close contact all the to coordinate their operations in order to ensure that everything is seamless so that no one even notices what is going on in the background of national and international finance, unless there is a breakdown as in the ECB excluding Greece from the payments system, effectively shutting down Greek banks.
Countries can also be shut out of the international payments system that is largely controlled at this point by the US. This is one thing the Iranian sanctions did. Similarly, many Russian individuals and firms have been shot out by sanctions against them. As a result Russia and China are setting up independent payments systems domestically and cooperatively internationally, which will include BRICS and other emerging nations that wish to avoid US control of their finances.
“However, this may have the effect of not increasing the money supply in circulation.”
The money supply that the private sector uses for transactions in the private sector is bank deposits, which are bank liabilities, and cash (currency in circulation that is a government liability). The size of currency in circulation is a function of the desire for nongovernment to hold cash rather than bank deposits.
It has nothing to do with the size of the monetary base other than the Fed ensures that banks always have sufficient required reserves on balance, lending settlement balances at the discount window at the penalty rate for overdraft.
Settlement balances are never in circulation, only cash, and cash in obtained by banks in exchange for settlement balances and by non-banks in exchange for deposit balances at banks.
Settlement balances exist only on the Fed’s spreadsheet and the Fed can always ensure that there are settlement balances sufficient to clear. Central banking is set up to do this as its major purpose. This avoids panics and bankruns.
The Fed adjusts the size of the monetary base using OMO in order to hit its target rate. The only point as which the Fed would have a possible issue with this is if it bought all the outstanding tsys, now about 12 trillion USD, which is somewhat far-fetched. But even then the Fed can also use other agency debt held by non-government. And the Fed can issue settlement balances I unlimited amount. So there is no conceivable circumstance in which the monetary base would not be able to handle settlement in the payments system.
A problem only arose during the recent financial crisis when banks were reluctant to lend to each other in the overnight market using the payments system because no one was sure who was still solvent. At that point, the Fed had to step in with QE1 and Congress with TARP (Troubled Asset Relief Program) to ensure that the financial system continued to work seamlessly.
Settlement balances are never in circulation, only cash, and cash in obtained by banks in exchange for settlement balances and by non-banks in exchange for deposit balances at banks.
I thought cash was only a small proportion of the dollars in circulation.
Bank loans create far more dollars than the Fed with deficit spending.
I pay my water bill, gas bill, my electric bill every month either by check or automatic electronic deposit from my checking account. So that money is in circulation. I never use cash (Federal Reserve Note bills). So, explain why those are not settlement balances. Some of the money in my checking account is monthly Social Security payments.
M1 money supply is bank deposits and currency in circulation as a cash (bills and coin).
The only way for the public to hold dollars is in cash. Bank deposits are simply entries on a banks books that are liabilities of the banks to customers, that is, promises to settle either in cash at the window or by honoring drafts on the customers’ deposit accounts. The drafts on deposit account never physically enter the private sector unless withdrawn as cash.
Otherwise, it is just shifting entries on deposit accounts around among different accounts within the same bank or among different banks, with settlement in settlement balances as required for interbank transfers after netting and for transactions with government like drafts to the Treasury for paying taxes.
The amount of bank deposits in M1 is not said to be ‘in circulation.” Only cash is said to be in circulation.
“How much U.S. currency is in circulation?
“There was approximately $1.37 trillion in circulation as of June 4, 2015, of which $1.32 trillion was in Federal Reserve notes.”
The rest is coin.
“I pay my water bill, gas bill, my electric bill every month either by check or automatic electronic deposit from my checking account. So that money is in circulation. I never use cash (Federal Reserve Note bills). So, explain why those are not settlement balances. Some of the money in my checking account is monthly Social Security payments.”
Your bank deposit is in liabilities of the banks, which the bank promises to convert into cash on demand at the window or into settlement balances to clear in the payments system as required after netting.
When the Treasury pays SS, the Fed credits your bank’s settlement balance account at the Fed with settlement balances (Fed liability-bank asset) by debiting the Treasury account (Treasury asset). The bank then credits your deposit account (bank liability-your asset).
Settlement balances are never in customer’s accounts at banks because customers’ account are on banks’ spreadsheets, and settlement balances exist only on the Fed’s spreadsheet. Settlement balances are not needed for transactions that involve intra-bank and inter-bank netting. Intra-bank netting takes place within the bank on its consolidated spreadsheet linking all its units, and inter-bank netting takes place in clearing houses.
What is not settled by netting is settled in the payments system in settlement balances that the bank promises to provide on the behalf of customers. Since banks don’t issue settlement balances ‚ only the Fed does, banks have to obtain them as needed to clear their obligations.
To understand this it is necessary to understand basic accounting and how various entries are marked up and down on different entities’ accounting records — Treasury, Fed, primary dealers, other banks, and bank customers in the cases we have been discussing. All of this is continually in balance in double entry bookkeeping so it is simple to track flows that increase and decrease stocks. This is called stock-flow consistency (SFC). It’s not possible to understand money & banking, finance and economics without understanding at the least the basics of this.
Tom, can the Fed simply lend settlement balance dollars to a bank under an agreement that the bank will use the securities it buys as collateral should it default on the loan: in other words, can the bank give the security itself to the Fed to settle the debt if the bank lacks the money to pay the Fed back after buying the security from the Treasury? Can the bank request further that there be no deadline in which the loan has to be repaid this way except on demand from the Fed?
For example, Fed loans full value of settlement balance dollars to a bank for buying deficit-spending securities from Treasury the bank has won the bid on. It comes to the auction with them and directly buys securities via the dealers who in this case are not first-buyers but just agents and facilitators of the auction who have taken the securities from the Treasury to sell for the Treasury. The dealers send the dollars received to the Treasury and give the securities to the bank. This is new settlement balance dollars so this is an addition to the money supply in circulation.
The Treasury spends the money on deficit appropriations. So, the money is gone into the economy. The Treasury does not have the money to pay back the bank for the securities, and the bank doesn’t have the money to pay its loan from the Fed. So, Treasury and bank easily reach a compromise to just perpetually roll over the securities with securities-swaps plus interest borrowed by the Treasury from banks given at the swaps to the bank. The bank gets free interest in the deal. It does not get back the principal (which it never had except as a loan from the Fed). The debts will never be repaid to bank nor to the Fed, so the swaps will go on forever…. unless, the Fed intervenes and asks the bank to give it the securities in return for canceling the debt to the Fed for borrowing the money for the purchase. (In any case, the deficit spending money is effectively debt-free money as well as new money.)
But the Fed will take the securities and swap them for new securities with the Treasury, allowing the Treasury to extinguish the obligation on the mature deficit securities by extinguishing the securities. This ends with canceling the debt obligation of the government to pay the holder its face value in settlement balance dollars.
I’ve been working on several scenarios in which Fed loans settlement balance dollars to various parties, such as the primary dealers, banks, under various arrangements, such as whether or not the dealers should sell securities to banks or to the Fed or simply mediate the sale without their possessing the securities themselves by buying them, or the Fed accept securities themselves as payment for the loans. There are many ways to skin a cat here.
But the key to having this work with dealers and banks is for the Fed to be the source of new money in the form of loans to the respective parties at the auction, so that the money given to the Treasury and spent on the deficit will be new money.
I now see why most people are ignorant of how this works, because it is not like anything ordinary folks encounter in their own experiences with personal or business loans. And it is complicated to explain.
But when the Treasury spends the settlement balances it has received for the securities for deficit spending, it spends them into the banking system, including the private banks. That money gets into these banks in various forms of deposits. These deposits get converted into cash (in part), because much of the work they do pays salaries, often in just cash. As an officer in the Army Reserve on active duty I served as the pay officer on a number of occasions for a battalion, and I had to pay with cash. Some of what we did, I’m sure, was paid for with deficit spending.
Tom, my conception of “circulation” must be different than the banking concept you mention, although there is an overlap. The banking concept would limit circulation to settlement balances, no? I would add to that the money created by bank loans and spent into the economy through cash purchases and interbank netting (as you put it). In other words, I am focused on a concept of ‘circulation’ that relates to inflation and deflation, and the flow concepts of inflows and outflows from a ‘sector of the economy’ called ‘circulation’. I call it that because it is money circulating from party to party as opposed to out of circulation, like out of the country via imports, in long term savings, in immature securities deposited in time deposits at the Fed, taxes not yet spent by government. Inflows are export dollars come home from sales of US goods to foreign buyers, government spending (from taxes and deficit spending), investments, bank loans and loans from nonbank financial institutions (like mortgage lenders). Hence, I think of the economy as a whole governed by the equation
ΔC = INFLO – OUTFLO
where C is quantity of money in circulation at any given time t, ΔC is the change in quantity of money in circulation at time t, INFLO is quantity of money flowing into circulation and OUTFLO is quantity of money flowing out of circulation at time t.
INFLO = ( E, G, I, L) = (exports, government spending, investment, bank loans)
OUTFLO = (M, T, S, P) = (imports, taxes, savings, payoff of bank loans).
BTW, the above equation is found in hydrology ΔS = INFLO – OUTFLO governing the change in quantity of water S (a stock) in a reservoir or swimming pool, and in physics in the first law of thermodynamics
ΔU = Q – W,
where ΔU is the change in the quantity U of internal energy of a system, Q is the amount of heat energy entering the system and W the amount of energy leaving the system as work. So, it is interesting that the same mathematical equation can be fundamental to economics (although applied to different stuff).
So, what is the quantity C of money flowing around between parties in a nation to which there are these inflows into it and from which these outflows flow out? It is crucial to the concept of inflation and deflation and a desired state for C, that is, C’, the level of money in ‘circulation’ or this system of interflowing money at which the economy is in full production and full employment with stable prices and wages. We need to link in our thinking this quantity C’ to inflows and outflows of money from this system. When C is less than C’, then we will have deflation. When C is greater than C’, we will have inflation. At C’ everything is ‘marvelous’, but it is a state that is not stable because the INFLO and OUTFLO are not stable. At C’ we wish
ΔC = INFLO – OUTFLO = 0, because we want to hold the system at that state C’. We seek to manipulate where possible the distinct inflows and outflows to increase or decrease C or to keep it at a specific level C’.
This is not a concept of neoclassical economics. I think it is Keynesian.
Note above, I include L (bank loans) and P (bank loan payments), which I have not seen done in MMT, which focuses just on the first three inflows (E,G,I) and first three outflows (M,T,S) but does not deal with the private banking sector’s contribution to money in circulation. Hence, my insistence on developing theory that includes bank lending and payments, because we have had inflations and bubbles not due to the Federal government spending creating inflation, but the private banking system’s lending and private borrowers’ spending doing so.
Isn’t government spending identical to taxes plus savings? I thought was a basic MMT tenet. Yes? No? It would simplify your inflow/outflow equations.
The deficit is government spending net of taxes. In the US, the deficit is offset by issuance of tsys in that amount. This drains the monetary base of the residual of settlement balances after taxes are withdrawn from it.
Since the liability for the tsys is on the side of government and the settlement balances are also on the side of government as a Fed liability, the aggregate net financial assets of non-government increase by the amount of the tsy issue, which is equal to the deficit as government spending net taxes. This flow adds to the stock of net financial wealth of non-government, that is, is net savings.
Saving is defined as the residual after spending (flow), so saving is a flow. Saving flows to the stock of wealth as net savings during the period.
It’s important to distinguish saving as flow that adds to the income statement form savings as a stock that adds to the balance sheet.
BTW, stock flow consistent analysis based on sectoral balances was worked out by Wynne Godley, formerly of the British Treasury and later econ prof. at Cambridge. He and Francis Cripps wrote an intro macro text published in 1983 as Macroeconomics. The definitive work is Wynne Godley and Marc Lavoie, Monetary Economics.
Tom Hickey, thank you very much for your response and for the references. I will make good use of them. My background is math/physics/engineering and system analysis and I tend to see these issues in a different way. For example when you equate savings to “flow” I think of savings as a function of time and the word net to me is an integral of a time function. Perhaps your references will enlighten me more in these areas.
Isn’t government spending identical to taxes plus savings? I thought was a basic MMT tenet. Yes? No? It would simplify your inflow/outflow equations.
Charles, government spending is an inflow. Taxes collected are an outflow. They cannot be placed on the same side of the equation. When the taxes are collected, they may or may not then be spent by Congress. When some of the tax money is spent, it becomes part of an inflow. If some of the taxes are kept and not spent, they remain an outflow in a surplus. Deficit spending is an inflow because it results in the creation of new money by the Fed that finds its way to the Treasury, which spends it into the economy.
What you may be thinking of are identities such as
Savings = Exports + Government spending + Investment + Bank Loans.
In other words, Aggregate savings later on equals the sum total of inflows at a given time.
As the money from the inflows on the right circulate in the economy, each buyer gets some of that in some form of income, spends some of it and keeps the rest. As the money goes around from buyer to seller, from employer to employee, each party may keep a portion of the original inflow that came in at a given time. Eventually after numerous cycles around and around, all of this inflow money that came in at time t ends up in aggregate savings and has been taken out of circulation.
This is why you need to constantly feed new inflows into the economy, because otherwise they will all end up somewhere in aggregate savings.
And then there is no money chasing goods and services, a total depression.
Stan, Re: spending/taxes/tsys sales.
The equation I was contemplating was: Federal Spending = Taxes + tsys sales. The flows are on the correct side of the equation. Enforcement of that rule prevents the federal government from adding to or deleting from the supply of money in the economy, leaving that power to the banking system. The problem is that taxing and spending are direct, obvious tools for control of the money supply while the banking system has that responsibility with very poor and weak tools; indirect control of interest rates and buying/selling government debt. I was reminded of this fact recently as the ECB intruded into the Greek government to control taxing and spending in an obvious attempt to control the money supply in the Greek economy.
Charles, your interest in “government spending = taxes + tsy sales” contains perhaps some confusion between taxes as collected and taxes as spent. If all of the taxes collected are spent and there is a balanced budget then government spending equals taxes collected. But if there is a surplus, then some tax money collected is not spent, and taxes collected minus government spending is positive. But if there is a deficit, then taxes collected minus government spending is negative; more is spent than taxes collected.
The equation above is for deficit spending. It could be written as
government spending = taxes collected + tsy sales.
Think also of taxes spent and taxes saved in a Treasury account. Taxes not spent are dollars saved. Dollars saved are kept out of circulation.
Ordinarily surpluses are not good for an economy, because a government that continuously runs a surplus can drain continuously money out of an economy, causing ultimately a depression.
But there are cases where surpluses do not hurt the economy. I believe Australia ran surpluses for several years, and were able to do that without deflation because Australian mining companies were selling tons of iron ore to China in return for Australian dollars. Exports can counter losses in government spending from excess taxes causing the surplus. That is why I do not think balancing the budget, fiscal balances, is what our government should aim to achieve. We need to attain a level of money in circulation that is sufficient to clear the market of a fully producing and employed economy at stable prices. And that may require deficit spending to counter negative effects of other losses from circulation through imports (dollars leaving the nation’s circulation to other countries, or into securities issued by the central bank of the buyer nation); or if the nation wishes to deficit spend to achieve homeland well-being, it may be encouraged to save and buy imports to draw out excess money beyond what is needed to attain full production and employment at stable prices and wages.
Stan, I would like to point out a rather fundamental point. A lot of the discussion here is about accounting and not about money. They are two very different things. Basically, counting money and keeping records of money is not the same thing as money. Confusing the two would be like confusing the box score from an athletic contest with the contest itself. They are simply not the same thing. I understand the need and value of accounting and of box scores but we should never confuse the records for the real thing. In the conversation here, Tom Hickey is obviously a true expert in accounting protocols throughout the financial system. He exhibits a level of understanding and clarity of explanation that I have never encountered before. However, to me, it is relevant but not the most important facet of a discussion about money.
Money has at least four functions. It serves as a unit of account, a medium of exchange, a vehicle for saving, and a means of deferred payment. State money also serves as a tax credit.
The MMT economist that has written the most on money is Randy Wray. Here are some of his papers on money, which are all available for free download at Levy Institute:
WORKING PAPER NO. 832 | February 2015
The Rise of Money and Class Society
WORKING PAPER NO. 821 | December 2014
WORKING PAPER NO. 778 | November 2013
Modern Money Theory 101
WORKING PAPER NO. 736 | November 2012
A Meme for Money View More
WORKING PAPER NO. 717 | May 2012
Introduction to an Alternative History of Money
WORKING PAPER NO. 647 | December 2010
WORKING PAPER NO. 512 | September 2007
WORKING PAPER NO. 459 | July 2006
Banking, Finance, and Money
WORKING PAPER NO. 438 | January 2006
Keynes’s Approach to Money
WORKING PAPER NO. 252 | September 1998
WORKING PAPER NO. 222 | January 1998
Money and Taxes
WORKING PAPER NO. 86 | March 1993
The Origins of Money and the Development of the Modern Financial System
Thanks for the link and reference list.
What you are calling money circulation is spendable money. That is the M1 money supply, which consists mostly of bank deposits. Bank deposits are created by bank lending and supplemented by net government spending.
It is quite difficult to get a handle on because of instantaneous credit creation through both credit cards and lines of credit that bring money into existence by being drawn upon.
Economist Steve Keen is now attempting to model this. You might want to follow what he is doing with a computer program called Minsky.
The formula that you are working with is the accounting identity
MV = (by identity) PQ (or PT)
that is called the equation of exchange.
While everyone agrees that MV = PQ is an accounting identity that always holds, there are different views of how to interpret it in economic theory. It is the basis of the quantity theory of money of monetarism. Keynesians have a different view of it. That’s much more than I can get into here.
Here is a post by John Harvey that explains the basics of the controversy over the equation of exchange from a Monetarist and also a Post Keynesian viewpoint.
While no one contests it as an accounting identity, there are issues with operationalizing the variables in an economic theory that imputes causation.
My equation ΔC = INFLO – OUTFLO, where C is quantity of money in “circulation” is a different aspect of it.
I think C itself is M, the quantity of money in circulation. So, the change in money ΔC in circulation is given as the difference between the inflows (of money) and outflows (of money). But the equation of exchange, I think, concerns how prices are set given a certain quantity of money in circulation and the velocity of that money moving through circulation. So, the level C’ at which there is full production and full employment at stable prices and wages will be determined by something involving P and T and M and V. I need now to study that equation, now that I know how it relates to C.
I suppose you now realize that I have come around to your views on settlement balances.
I don’t know all the details about how bank reserves are recorded at the Fed, or what the liberties and constraints are on the Fed in buying and selling things. That’s why I’m asking you about how the Fed may give settlement balances to banks (instead of lending it to them), or how it lends, how it agrees to when the borrower has to pay it back, etc.. Can the Fed accept the collateral securities as payback for a loan for purchasing them? But I hope I am correct in thinking that the Fed will lend the full amount of the principal to a bank with which to buy securities at the Public Auction, using the securities as collateral. It is essential that newly created money from the Fed is used to buy deficit-spending securities. That assures that the money spent in deficit spending will add to the money supply. That can come from overdraft loans from Fed to primary dealers of money with which to buy the securities, or loans to banks at “the discount window (?)”.
But there are many ways the settlement balance dollars can end up in deficit spending. Sometimes it can come from primary dealers (who got it from the Fed in overdraft privileges) and from the banks (who got it by borrowing from the Fed at the discount window (or wherever). The banks want it because they get debt-free interest, and the principal (they lose) didn’t really cost them anything to get the loan, other than agreeing to yield up the securities if the banks can’t come up with other money with which to pay back the loan.
BTW I have read Steve Keen’s Debunking Economics 1 1/2 times. It’s a great book!
How do foreign investors with dollars earned from selling us imports get their dollars turned into settlement balance dollars with which to buy US securities?
For that matter, how does anyone with only bank-created money, get settlement balance dollars with which to buy securities or pay taxes?
Suppose it is some foreigner who has come into the US with little or no dollars who buys a lottery ticket in some QuickTrip store and wins the lottery. Will he be given settlement balance dollars as his winnings? If not, how could he invest these dollars in US securities, if they are only bank dollars?
Stan, looking at that equation MV = PQ or PT: the left side has units of $/unit time so to be a sensible equation the right side must have the same dimensions. P is $ so that index number must have the dimension of “per unit time.” Is that how those indices are defined?
Charles3000, I am not actually working with the exchange equation. Tom thought I was. I recognize that given a base of M dollars in circulation, then the level at which we could have full employment and production at stable prices might be a function also of V, and that would have some bearing on prices. Beyond that I haven’t looked deeper into that equation.
But V I suppose is number of transactions (on average) per unit time. Prices on the right side would be an average for whatever interval was estimated, and Q or T, whichever, must be per unit time also.
I’m actually not an economist by profession but a quantitative psychologist, retired. Since the crash in 2007 I’ve become interested in issues of the national debt (I don’t believe it is a problem for taxpayers), MMT and how you have to consider exports/imports and Investments/savings along with Government spending/Taxes. The differential equation ΔC = IF – OF describes the change in quantity of money in circulation C at a given point or interval in time, with IF inflows into circulation of various sources of money and OF outflows of various sources from circulation. Instead of focusing on balancing the budget by making Govt spending G = T tax revenues collected, we seek to have the quantity of money in circulation be a value C’ at which there could be/is full production and full employment at stable prices. At that point we would seek to balance IF = OF, or IF – OF = 0, to force no change. (Keep a good thing going). That would involve a careful analysis of sources of inflows and outflows and seeking to find where we might have control of some of these sources to achieve the balance of IF – OF = 0. But once you see this is what govt should be concerned with and not just balancing the budget G = T, you see the big picture of how to optimize the economy for everyone’s benefit.
I’m convinced that with imports draining money out of circulation in the nation, Bush-Cheney were able to fight the war in Iraq with deficit spending but no inflation. Cheney was right, deficits don’t matter (since they are inflows of new money into circulation and you can counter their inflationary effect by outflows of imports). The large increase in the national debt was not due to government spending, but to imports from China and Japan and other countries being converted into their purchases of securities and that money being taken out of circulation and not becoming a problem for deficit spending on the war in Iraq. In 2010 87% of the public (national debt less value of intragovernmental debt) debt was to investors whose money was still in our Federal Reserve bank as time-deposits, ready to be returned to the investors on demand when the securities matured. And since imports have continued to drain dollars out of circulation and were converted by the importers into securities, these securities now make up near 90% of the national debt figure. The reason the national debt is not a problem (for taxpayers) is because it is (1) being rolled-over by the Treasury (for deficit spending), (2) is backed by dollars in time-deposit accounts at the Fed, readily available for return to the creditor with interest (created out of thin air by the Fed), (3) is actually being dismantled by the Fed’s Quantitative Easing in which it purchases all mature US securities held by banks, and that includes deficit spending securities. Fed will swap its mature (deficit spending) and other securities for new ones. That ends that national debt, while the new securities are just held in custody by the Fed until they can be used to sell them to investors and banks to drain money from circulation or bank reserves to fight a rise in inflation. Treasury will extinguish those mature securities it gets back. Taxpayers don’t have anything to do with redeeming the debt.
Stan, I recommend Frank N. Newman’s “Freedom from National Debt” as an interesting and informative read. Regarding the relationship between prices, money supply and employment, there is at least one other important variable, resources and an important element of resources is available labor. I’ll write more, Stan, when I have more time.
You mean defining I, G, X, L, M, S, T, P as at per unit time? Yes. Flows are quantities flowing past a meter during a unit of time.
At this point, you need to start thinking in accounting terms, that is, corresponding entries on the parties to a transaction’s income statement and balance sheet. Financial aggregates are put together based on that.
There is a strict separation between the government and non-government books. The Treasury issues tsys as its liabilities appearing on its balance sheet, and the Fed issues settlement balances as its liabilities. Settlement balances exist only on the Fed spreadsheet,
Banks’ books exist only in non-government, as do non-bank entities books. Settlement balances enter non-government only when they are exchanged for settlement balances at the Fed for vault cash, which is then distributed to bank customers as demanded.
Banks have accounts at the Fed and this exists as settlement balances in the payments system.
The only way that banks can get settlement balances is either 1) from Treasury directly the Fed to credit settlement balances to banks that act as agents for recipients of government spending and transfers (like SS) or 2) from Fed lending. I have explained this in some detail.
Banks have to get settlement balances either from government spending that credits their accounts at the Fed in the payments system, or by borrowing from Treasury, or by settlement of drafts in the payments system or by borrowing from other banks that have an excess they seek to lend out in the overnight market.
The Fed always ensures that there are sufficient settlement balances in the payments system to clear the system while allowing the Fed to hit its target rate. It does this chiefly using OMO in normal time. Now the Fed is setting the rate by paying interest on excess settlement balances, so there are more settlement balances in the system than needed since the interest rate is not affected under these conditions.
All this is done on the side of government, excepting exchanging settlement balances for cash, by accounting entries “out of thin air.” The Fed just prints up bills to meet demand for cash.
Banking is straight forward, as I have explained. Banks issue loans to customers as bank assets with a corresponding credit to the customers’ deposit accounts, which constitutes a bank liability (the customer is loaning to the bank). The customer has an asset in the deposit that the bank owes and a liability as the loan from the bank that has to be repaid with interest.
Bank loans have to be negotiated. Loans may be made for specific reasons, such as firm capital or housing purchases. Many loans are pre-approved and my be used as desired. These are lines of credit, such as one’s credit card.
Again, all this occurs in non-government in terms of accounting entries, other than spot transactions in cash.
All the movement of funds around the economy is in terms of crediting and debiting different accounts that correspond to the myriad transactions that occur everyday. It’s designed to be seamless.
Accounting reports are generated from this as data and aggregates are computed by combining reports.
This is the basic structure that one needs to understand. Then one needs to access the specific data that one needs for the questions that one poses. That involves accessing collections of data that are available. In the US there is lots of relatively up to date data available. This is what financial analysts and economists work with.
For example, what you are calling C is M1, that is liquid funds that are immediately spendable. To see the data on M1, just go to FRED at the FRBSL
Tom, while your answer above was very informative, it dodged answering directly my specific questions about how banks might get settlement balances as loans from the Fed (with no specific repayment date) for purchase of US securities. In my view deficit spending must always be in new settlement balances created by the Fed out of thin air at the time in a direct path to the purchase. Otherwise, ‘old’ settlement balances, already in circulation from earlier deficit spending and used by, say, banks to buy the securities from the Treasury will not add to the money supply, because it will withdraw settlement balances from circulation and then spend them back into circulation with a net change for the quantity of settlement balances in circulation of zero. So, please give me an answer. How can the Fed create new settlement balances (out of thin air) and have them lent or given to whomever will buy the securities from the Treasury?
What you suggest below does not seem to include the case I am considering:
The only way that banks can get settlement balances is either 1) from Treasury directly [?] the Fed to credit settlement balances to banks that act as agents for recipients of government spending and transfers (like SS) or 2) from Fed lending. I have explained this in some detail.
Perhaps I don’t understand your terminology, e.g. ‘payments system’.
And why not simply also ‘borrowing from the Fed’? I don’t see what constrains the Fed from creating settlement balances and using them in making loans of the purchase money to banks for the purchase of US securities, especially if it makes such loans to primary dealers for the same purpose through their use of overdraft privileges at the Fed.
If banks aren’t doing this, they should be clamoring it be done, because otherwise they are getting screwed and the economy is also, because deficit spending will not increase the money supply in circulation.
Banks want to earn interest on creation of new money. They had laws passed that favor them in this (1917). But they would not want to collect settlement balance dollars from circulation as fees, cash payments of debts by bank clients, etc. and then buy the securities with this money. The reason is that the Treasury will never have sufficient tax revenues on hand to cover current deficit spending and previous deficit spending (to redeem the prior debt), it will have to perpetually roll-over the debt. That means the banks under that arrangement would never get back their principal. And that is hard money.
If immediately prior to purchasing the securities the banks get their settlement balances from the Fed as a loan with no set redemption date and a promise to the Fed to use the securities to be purchased as collateral, then the Fed can know that it can, at any time it chooses, demand surrender of the securities to it to settle the debt. But it can allow the Treasury and banks to continue with their roll-overs at each of which the banks earn new interest, until some higher purpose requires the Fed’s intervention to acquire the securities, e.g. Quantitative Easing. This money is ‘soft money’ for the banks. They don’t have to work to get it. They can give up the principal as something they never had in the first place, so nothing is lost, as long as they get their interest.
There may be benefits in allowing them to get debt-free interest. It allows them to put more money in circulation than what they loan and require interest payments on. (Deficit spending with newly created money also adds to the money supply, so there will likely always be enough money in circulation for private bank borrowers to pay back the principal and the interest on their loans). There may even be a margin in which they can require higher interest on their loans than would be otherwise available in circulation.
Because the funds used in spending on consumption and investment are mostly created endogenously through lending, it is difficult to define M the equation of exchange operationally. M1 is an approximation but funding for purchases is added to by tapping credit lines, e.g., credit cards. There is also a lot of non-bank lending. Most vehicles are now financed not by banks as before but they the finance arms of the vehicle manufacturers. Other than a down payment, other funding is not required until loan payments come due, often only after a 6-mo grace period as a sales teaser.
M1 is a good rough guide but in a hot economy with lots of sources of credit and easy credit at that, there may be a lot more income being drawn forward than is indicated by M1. For example, in a hot economy credit card lines of credit may be increasingly drawn down. In addition, vehicle sales in the US are now running at about 16 million units a year, and most may be non-bank financed.
“Tom, while your answer above was very informative, it dodged answering directly my specific questions about how banks might get settlement balances as loans from the Fed (with no specific repayment date) for purchase of US securities. In my view deficit spending must always be in new settlement balances created by the Fed out of thin air at the time in a direct path to the purchase. Otherwise, ‘old’ settlement balances, already in circulation from earlier deficit spending and used by, say, banks to buy the securities from the Treasury will not add to the money supply, because it will withdraw settlement balances from circulation and then spend them back into circulation with a net change for the quantity of settlement balances in circulation of zero. So, please give me an answer. How can the Fed create new settlement balances (out of thin air) and have them lent or given to whomever will buy the securities from the Treasury?”
Bank can always borrow from the Fed using repo by tsys they own as collateral. They can borrow existing settlement balances in the payments system from other banks in the overnight market. The Fed makes sure that there are sufficient settlement banks using OMO. Or the bank can run an overnight overdraft at the Fed using the discount window.
I don’t see what the issue is here. The Fed continually creates settlement balances in the payment system as needed through OMO and reduces them by reverse repo. There is a huge excess of settlement balances from QE through the Fed purchasing tsys for its own balance sheet (asset) by issuing settlement balances, which its own liabilities.
There is never an issue of lack of settlement balances because the Fed controls the amount of them available in the system through its operations. The Fed never gives anyone settlement balances. They enter the system either by Treasury spending, Fed lending, or the Fed purchasing tys from non-government for settlement balances.
The only issue that arises is when the Treasury has reached the debt limit and cannot issue more tsys to get the credits to its account for spending by directing the Fed to credit bank accounts. The Fed cannot give them to the Treasury or lend them either. This is why either the 14th Amendment or some gambit like the platinum proof coin is needed to circumvent the debt ceiling by allowing the Treasury to meet existing government obligations without issuing more tsys.
“In my view deficit spending must always be in new settlement balances created by the Fed out of thin air at the time in a direct path to the purchase.”
Deficit spending is spending out of the Treasury account at the Fed in excess of tax receipts. The Treasury gets those credits from auctions of tsys that are conducted for it by the Fed as agent. The tsys are purchased with existing settlement balances in the system (remember the Fed can always increase the amount of existing settlement balances by issuing them against tsys through OMO or POMO) or Fed lending to the dealers using the issue as collateral.
As far as I can see you are creating a problem where there is none. You seem to be thinking of a fixed stock of existing settlement balances when the stock is being constantly adjusted by the Fed to fit changing conditions by lending and repoing and reverse repoing tsys. which the Fed can easily do since it issues settlement balances.
The Fed can always increase the monetary base with “new settlement balances” (newly issued settlement balances) by buying existing tsys in the market, for instance. The dealers can then borrow in the overnight market to obtain the settlement balances to settle if the dealers were the not the providers of the existing securities.
The system is set up to clear and the Fed ensures that it does while also managing the policy rate if it not setting the rate to zero or paying IOR. This is the day to day operations of the Fed.
“Perhaps I don’t understand your terminology, e.g. ‘payments system’.”
The payments system is the spreadsheet run by the Fed in which final settlement between government and non-government, and among banks in which transactions are settled after netting. This takes place only in settlement balance that exist solely on the this spreadsheet, which the Fed controls as the US government’s currency issuer.
Banks can’t issue dollars. They can create deposits denominated in dollars that they promise to exchange for dollars to settle, either as cash at the window or in the Fed’s payments system after netting.
Deficit spending leaves the amount of settlement balances unchanged after non-government accounts are credit and the tsys auctions are settled. What increases is the amount of tsys auctioned, the settlement balances from which credited the Treasury. The spending returns the settlement balance to the monetary base leaving it unchanged.
The flow of settlement balance banks to the Fed at the tsys auction and from the corresponding credits for Treasury spending to non-government accounts. This is a wash — net zero — in the payments system. The monetary base is unchanged. The settlement balances get drained into the newly issued tsys, which the Fed can at any time convert back to settlement balance in the payments system by repo or purchase of tsys in the market.
The stock I see is those settlement balances that were spent by the treasury in previous months or years or spending cycles. Now, you have said that settlement balances cannot circulate. But is the Treasury’s account you mention above from which it spends not filled with settlement balance dollars? What does it spend if not those? If the deficit is for infrastructure spending, then in what form of money does Treasury spend on that?
“Money in existence” is not the same for me as “money in circulation”. Much of the money floating around the auction is in existence but not yet in circulation chasing goods and services. That’s what the equation ΔC = INFLO – OUTFLO is about. For government spending to involve new money, the deficit has to be new money spent into circulation. The money in government spending from taxes is old money extracted from circulation.
Government spending = spending old tax revenues + deficit spending.
If, however, deficit spending is based on ‘old money’ taken from circulation, then that is not new money added to the money supply, but just a movement of old money from one set of hands to another set of hands without change in the money supply in circulation. New money has to be newly created, by the Fed and spent into the economy. The Treasury in spending on infrastructure will write checks to consortiums of private construction companies, who will divide up the money according to rules set by the government, among their members. Or large companies will bid on the jobs to be done, perhaps even proposing what would be done. Then government decides to whom to write checks to do that, whatever. The money coming to these companies are not cash. It would take an army of armored money carriers to deliver it to the companies’ banks. No, a simple electronic check could transfer that money to the bank accounts of the companies. So, in what form is the money? It is circulating, chasing goods and services.
I have already explained this several time. When the Treasury spends it directs the Fed to credit recipients’ accounts at banks. Then the bank has a credit at the Fed as an asset and the Treasury’s account (asset) is debited. The bank has an asset in the from settlement balances (Fed liability).
The bank then credits the recipients’ deposit accounts in that amount (bank liability, customer asset). If the customer wants cash at the window, the bank reduces its vault cash (bank asset, government liability) to exchange its liability for the government liability. The bank then exchanges the settlement balances it received from the Fed crediting its settlement balance for the Treasury spending into cash to replenish its vault cash.
Say the customer writes a check to another bank that requires settlement in the payments system instead of through netting. Then the bank’s settlement balances is this amount is transferred to the settlement balances account of the recipient of the check. That bank now has the settlement balances in its account and it credits the person’s account that cashed the check.
The transactions that you handled as paymaster took place within government among agencies of government using the Fed as the government’s fiscal agent. Non-government was not involved and the accounting was on the books of different government agencies.
The problem with trying to explain something to someone else, you have to figure out how to put it in terms that the other person will understand. That means you have to read what the person is saying and figure out what he/she is thinking. Not easy to do.
What is not clear in my mind is the places in the system where bank money can be exchanged for settlement balances. I understand that cash is a form of settlement balances. But, am I wrong when I think of ‘cash’ as bills and metal money? But cash is very inconvenient to pass around when it comes to large amounts of dollars.
So, I am thinking how infrastructure spending by government, which will involve a lot of deficit spending, will begin as settlement balances provided to the Treasury from the sale of its securities at the auction. Then it will be put in the Treasury’s general account at the Fed. Soon thereafter the Fed will draw some of that settlement balance dollars in the form of checks drawn from that account and send the checks to those companies that won the bids for infrastructure construction. Some of that money will be used to buy equipment (dump trucks, road scrapers, rollers, materials (cement, road tar, steel bars, sand, gravel, etc..). Other portions of the money will pay wages. At some point cash will be disbursed, although the workers might spend whatever they have received for wages with credit cards. No cash will be used.
So, what is the form of dollars at each step along the way, and where does it change from settlement balance dollars to cash or even bank dollars?
And when there is an exchange of forms of money, where does the bank dollars go that were swapped for cash? Or where do settlement balances get exchanged into bank dollars?
And is the interface between government and private sectors limited in unreasonable ways that is very inefficient?
This is all based on accounting. All transactions appear on the respective parties books as entries. The Treasury has its book, so does the Fed. Both issue reports for public consumption. When the Fed issues tsys, it is a liability. When the sale takes place and the Treasury General Account is credited it is an asset. Settlement balances and FR notes are Fed liabilities, and the tsys that the Fed buys to transfer settlement balances to the payments system/monetary base are Fed assets.
Every bank has its own book, as does every firm. Most and more household are actually keeping books on software, but most people have some way of keeping track of their checking accounts by recording deposits and drafts against deposits. All transactions in the economy takes place by crediting and debiting various accounts or using cash for spot transactions.
The interface between government and nongovernment occurs in the payments system on the Fed’s spreadsheet where banks that are members of the Fed have settlement balance accounts and where they can exchange settlement balances for cash.
When government spends, a bank’s settlement balances account is market up on behalf of the recipient and the bank credits the customer’s deposit account on its own books. That is spendable funds as part of the M1 money stock.
Banks extending loans also creates deposits that become part of the M1 money stock.
The M1 money stock is used for transactions other than spot settled in cash.
Settlement takes place first by intra-and inter-bank netting and then final settlement takes place at the Fed in the payments system using settlement balances in banks’ accounts there.
This is a seamless system that is highly efficient — now it is fully digitized. And owing to the Fed’s lender of last resort function (member banks can run overdrafts) and the FDIC’s deposit insurance, the system is also very stable.
“The stock I see is those settlement balances that were spent by the treasury in previous months or years or spending cycles. Now, you have said that settlement balances cannot circulate. But is the Treasury’s account you mention above from which it spends not filled with settlement balance dollars? What does it spend if not those? If the deficit is for infrastructure spending, then in what form of money does Treasury spend on that?”
Settlement balances are used only for settlement among members of the Federal Reserve System among each other and with the government (after netting).
Settlement balances are part of the monetary base and the Treasury account is not counted in the monetary base.
When the Treasury spends, it directs the Fed to credit banks’ settlement balance accounts and banks in turn credit their customers deposit accounts. This flow of spending increases the monetary base (stock). The settlement balances are Fed liabilities used to settle taxes, so they are tax credits for nongovernment.
When taxes are collected, bank customers send drafts to the Treasury that are cleared in the payments system. Banks debit their customers’ deposit accounts and the Fed debits the banks settlement balances account. This withdraws the settlement balances from the money base and this flow reduces the monetary base.
When the Treasury spends, currency in the form of settlement balances is added to the system as the Fed credits banks’ settlement balances accounts, which the banks hold as an asset and can use for settlement in the payments system or exchange for vault cash.
When the Treasury receives taxes, the settlement balances are canceled since the credits the Treasury receives are not counted toward the monetary base and no longer exist in the payments system.
(TT&L accounts just delay the withdrawal of to settlement balances from the monetary base/payments system as a means that the Fed uses to adjust the size of the monetary base, that is, the liquidity in the payments system. It does this as operation like OMO to hit its target policy rate when it is not setting the rate to zero and not paying IOR.)
When Treasury spends more than it taxes, then it deficit spends using tsys it issues and directs the Fed to auction to the dealers for settlement balances. This drains the money base/payments system of the amount of government spending in excess of taxes. Thus the amount of the deficit is equal to the amount of newly issued tsys.
This drains the monetary base of the amount of settlement balances injected by spending. This is a means of reducing the size of the monetary base so that the Fed can use less OMO or POMO in adjusting the size of the monetary base to hit its targeted policy rate.
Spending creates currency (Fed liability) in nongovernment (banks settlement balance accounts and vault cash) and taxation withdraws it. Issuance of government securities shifts currency (banks settlement balances) to tys.
Currency as settlement balances and currency held as cash as vault cash or as currency in circulation is a Fed liability. Government securities are liabilities of the Treasury. Thus, the amount of settlement balances, cash and tsys by nongovernment constitutes the aggregate net financial assets of nongovernment.
When the Fed uses OMO or POMO to shift between settlement balances in the payments system and tsys, it is only shifting the composition of nongovernment net financial assets between zero maturity and non-zero maturity, but not the amount held.
From the MMT POV currency is a tax credit, and tsys issuance is an interest rate setting operation. The stock of tsys held by nongovernment is the stock of tax credits not yet used to pay taxes.
“So, in what form is the money? It is circulating, chasing goods and services.”
When the Treasury spends, that is, pays invoices incurred by other agencies in accordance with budgetary appropriations, makes transfers like SS, and pays interest on tsys, it directs the Fed to credit customer accounts, which the Fed does by crediting the settlement balances accounts of banks of the recipient and debits the Treasury account, which has credits from taxes and sale of tsys.
Banks then credit the recipients’ deposit accounts for spending into the economy or saving.
Banks also issue loans that create deposits, which are also used for spending.
Spending takes place mostly though bank drafts, use credit cards, etc, that are settled first by netting and then in the payments system using the settlement balances that the Fed makes available for liquidity using its monetary operations to shift back and forth between settlement balances and tsys to adjust the size of the monetary base when it is not setting the policy rate to zero or paying IOR. Now the Fed is paying IOR and the monetary base is irrelevant for hitting the target rate.
Tom said in: reply to Stan’s “So, in what form is the money? It is circulating, chasing goods and services.”
This seems close to what I’m trying to understand. In the above you seem to say that the Treasury receives settlement balances for its securities it has sold to banks. It then turns around and spends that into the economy (via the banking system) by crediting the settlement balance accounts of banks receiving it and debiting the Treasury’s account which has credits from taxes and sale of tsys. In other words, Treasury spends with tax money plus money it got from sale of the securities.
Does that lead to what seems to be a transformation of federal settlement balances into bank dollars…. ???
How does that transformation take place? Does the bank just match its federal settlement balance account gain with new bank-created entries in bank dollars netting with the settlement balance gain? There seems to be an exchange here.
My next question is, if this exchange of settlement balances for bank dollars takes place here, what happens to the settlement balance accounts of the banks? Are they debited as the bank dollars are created and added to the customer accounts? The settlement balances vanish into thin air and bank dollars in net value appear in the customer accounts. In other words there is no physical transfer of settlement balances into bank dollars, no physical exchange, but just writing into a spreadsheet what equals in sum in the settlement balance accounts.
I’m just guessing here in what seems to me the most plausible way this could be done.
Nevertheless, you have said that settlement balances (except for cash) never circulate chasing goods and services as liquid money. So, I have to conclude here that there has been a transformation in some sense of federal settlement balances into bank dollars, which do circulate. Again how?
What you have written is quite informative, but still somewhat abstract. This is why I am asking you these questions, to obtain clarification.
Tom, I know I am frustrating with my questions. You think you are answering them, but not fully. I need feedback on my particular interpretation of the question of transformations from settlement balance to bank dollars. What you describe seems to do that, with the separate books and spreadsheets.
So, am I interpreting you correctly?
I gather that banks have settlement balance accounts at the Fed. How and where they get them, and what goes into them is not fully clear. But I will grant that they exist.
Now you say Treasury spends by drawing down on its Treasury account at the Fed and depositing settlement balances in the banks settlement balance accounts at the Fed. Treasury also directs that the banks draw from these accounts to allocate various customer accounts funds from the settlement balance accounts. Actually no physical withdrawal takes place. There is just a debiting of the settlement balance accounts and a crediting of bank customer accounts in equal increments. The accounts are in different forms of dollars, Fed dollars and bank dollars. The banks have access to their settlement balance accounts at the Fed in the form, say, of copies??? So, they know also what has to be decremented and to be incremented in the bank’s books. It looks like a transfer across a divide, but we didn’t see anything in the interim going across a divide. It’s almost like quantum mechanics. It may be due to an intervening person working at a computer spreadsheet, digitizing entries in various columns.
A quote from Stan: “So, I don’t see what your $1T coin would accomplish. ”
The T$ coin is a threat to the status quo. It is subtle but very real. The US has twin monetary systems running in parallel. One is paper currency and the banking system enjoys the seigniorage on paper currency except for the small refunds demanded by congress. The other is coin and the federal government takes full seigniorage on coins, both positive and negative (as on pennies and nickels) and the T$ coin would carry a very high seigniorage. It is sometimes identified as HSC. If all deficits were funded via HSCs then at the end of 30 years, the oldest tsys sold, the national debt would be retired and the Fed would be left as a clearing house only with no power to alter the quantity of money in circulation. That power would then revert to where it should reside, with the congress, using taxing and spending to control the money supply, the logical tools for that task.
L. Randall Wray in working paper 792 quotes the following:” Ingham argues “There is, as Amato and Fantacci succinctly observe, ‘no liquidity without a lender of last resort, no lender of last resort without an irredeemable consolidated debt, and no irredeemable debt without an immortal state’” (Amato and Fantacci 2012: 236; quoted in Ingham 2013b, p. 23).” The “irredeemable consolidated debt” is the national debt which would be ended eventually if the T$ coins are used to finance deficits.
I appreciate Tom Hickey providing me with a link to Wray’s working papers. They are very informative.
This is fine for deficit spending securities. But if you buy back all those investor securities, which were not used to fund government projects, you release all those dollars into circulation unless the investors reinvest them in some new securities or something outside the bounds of circulation in the US system. So, we need to be precise over what securities are the real problem (deficit-spending securities). The investor securities are backed by their time deposit feature in the Fed securities reserve account.
I know the debt-ceiling law does not make this distinction and needs to be changed.
The key is understanding that all that is happening is crediting and debiting various accounts on different parties books in the unit of account in a currency zone. The USD zone is larger than just the US, for example, and includes everywhere in which USD are in use.
Some entities have access to the central bank and some don’t. Those that don’t have to go through those that do to deal with government or for final settlement.
Those entities that have access to the Fed have accounts at the Fed on the Fed’s books. Those entries include assets in the form of settlement balances and liabilities in the form of loans from the Fed.
For example, when the Treasury spends by directing the Fed to credit the recipient’s account, the Fed credits the recipient’s bank’s account at the Fed in settlement balance. The Fed debits the Treasury General Account (TGA) and credits the bank’s account with settlement balances that increase the monetary base. So the TGA loses an asset and the bank gains an asset.
But the bank has a liability to the recipient, who has a deposit account at the bank. So the bank credits the customer’s deposit account, which is a bank liability. When the customer draws down the account and the bank settlement the drafts, it uses the amount of settlement balances it received from the Treasury. Thus the bank debits its settlement balance account (asset) and credits the customer’s deposit account, erasing the bank’s liability. The recipient of whomever the customer made the draft to now has the asset, e.g, cash if the customers withdrew cash to settle or another bank if the the draft required transfer of settlement balances in the payments system.
So the answer to your question of how the settlement balances in banks’ accounts at the Fed get into customer’s deposit accounts for spending is just crediting and debiting the relevant accounts so that everything stays in balance. The only time that is a transfer of money tokens is when banks exchange settlement balances for vault cash and when they either pass vault cash through the window to customers’ drawing down deposit account or from customers’ depositing cash. Otherwise, it is all digital entries that record various transactions in a unit of account and between units of account in international transactions.
But this is only the way the US system works because it has a reserve requirement that requires banks to hold settlement balances at the Fed as a percentage of lending. Canada has no such requirement, for example.
Brian Romanchuk has written a book on this and blogs about it.
Stan, here is link to a short comment on how the system works in the UK by Neil Wilson.
“Does that lead to what seems to be a transformation of federal settlement balances into bank dollars…. ???
How does that transformation take place? Does the bank just match its federal settlement balance account gain with new bank-created entries in bank dollars netting with the settlement balance gain? There seems to be an exchange here.”
Yes, the settlement balances just get shifted from Treasury spending into the economy into Treasury securities. In this way the amount of settlement balances remains the same.
If the Treasuries would be bought by customers paying through bank deposit accounts, then the amount in deposit accounts would remain unchanged and M1 would remain the same afterward, but usually there is a mix of buying of Treasuries among auction subscribers that includes banks and non-banks
When a bank customer buys a tsys through a bank, the bank purchases the tsys with settlement balances through the Fed payments system and debits the customer’s deposit account at the bank.
When the Treasury directs the Fed to credit a non-bank recipient for its spending, then the Fed credits the recipient’s bank with settlement balances in the payments system and the bank credits the customer’s deposit account.
The bank uses the settlement balances to settle drafts on the customer’s account as needed, or to replenish vault cash if the customer withdraws cash at the window.
The transactions with government and nongovernment and among banks takes place in settlement balances at the Fed in the payments system.
Transactions in the economy take place using drafts on bank accounts and cash in circulation withdrawn from banks.
This all happens through double entry crediting and debiting of accounting records so that all books are always in balance and reflect the flow of funds in terms of sources and uses.
Tom, I digress a bit here to earlier statements by you of the need for the Treasury’s platinum coin (say in denomination of $1Trillion) which Treasury would use to cancel the national debt.
Just what securities would you buy with settlement balances exchanged for such a coin? Would you buy all those investors’ securities? They make up almost 90% of the national debt. But what would you do with all the dollars in their time-deposit accounts? And they are not there for deficit spending, but as investments. No government program is funded by them. They exist as a way to keep their dollars out of circulation. Otherwise if they enter circulation again, they might cause inflation. We would have to cut back on some valuable government programs into infrastructure, education health, defense.
Only deficit-spending securities fund government programs.
Tax dollars fund the rest of the costs of these programs. But are you sure there are active securities now in existence that require even a $1T coin to cover? I think the Fed with its QE has quietly bought up most of the mature securities used to fund deficit spending. Look at current figures for the amount of securities held by the banks. And it has swapped these for new securities with the Treasury. That extinguished the deficit-spending securities. What remains is something on the order of $500 Billion which represent the latest years’ deficit spending plus interest. At this time the Fed holds about $2.5T in securities, but these are ‘inactive’, not held by any nonfederal entity who will demand payment in cash when they mature. They are like dollars stored in a vault somewhere for later access. The Fed cannot sell them to the Treasury, nor it lend to the Treasury. It can only swap securities for securities with the Treasury.
The problem with the debt ceiling is legal. It is unconstitutional. Congress cannot pass laws that limit or constrain in general its absolute powers as given in the Constitution. It has absolute powers to pay the debts and to borrow. The administration–someone qualified–needs to sue Congress for passing an unconstitutional law, the debt-ceiling law. Maybe currently the reason the administration has not sued in the Supreme Court is because they think the ignorance of the GOP justices–we know who they are–would declare this legal. But the issue needs to be talked up by the administration in the media. Can we get the GOP in Congress to recognize the illegality and unconstitutionality of it?
So, I don’t see what your $1T coin would accomplish. It would mess up an excellent way of draining dollars out of circulation so that the government can deficit spend on other programs. (I refer to time-deposit sequestration of dollars from investors who buy US securities).
So the Fed gained new securities, but at this point is is acting just as a custodian for them. There is no nonfederal entity holding them. The Fed will sell them to investors and banks if inflation occurs to drain dollars out of circulation and constrain bank lending by draining the banks’ reserves. Otherwise if they mature, the Fed will turn around and swap them again for new securities. It needs new securities to sell.
The purpose of the platinum proof coin when it was brought up at the time of the debt ceiling kerfuffle was to increase the balance in the Treasury General Account to fund US government obligations already existing under current appropriations to avoid a default.
Those who proposed this gambit said also that the platinum proof coin could also be used to pay down all the existing tsys available on the market. This was not a serious policy proposal, but simply a statement that the US can always not only fund itself by direct issuance under current law but also redeem its entire debt.
Could this result in inflation? Only if supply doesn’t meet increased existing demand that might expand, that is, the limit if use of all available resources is reached with effective demand still rising continuously. Then inflation would result. Under functional finance inflation could be addressed effectively by increasing taxes.
“So, I don’t see what your $1T coin would accomplish. It would mess up an excellent way of draining dollars out of circulation so that the government can deficit spend on other programs. (I refer to time-deposit sequestration of dollars from investors who buy US securities).”
As I have said previously, sequestration of deficit spending in tays does not decrease spending power because tsys are highly liquid in the market, the Fed always ensures the market clears and tsys are also used as collateral for borrowing. If tsys were non-negotiable and could only be held to maturity would they neutralize deficit spending.
IN addition, in an endogenous monetary system where bank borrowing constitutes the bulk of spendable funds in the economy, and the central bank conducts operations so that all transactions clear, the central bank simply purchases tsys for its book to add settlement balances needed. So the amount of tsys held by non-government doesn’t include all tsys issues. The Fed holds trillions of tsys on its balance sheet as assets, for example.
I think you misunderstood my comment above, Tom. I was not referring to sequestration of deficit spending, but the sequestration of investor dollars in time-doposits at the Fed corresponding to the securities. Investors buy US securities like you or I might buy a CD at a bank to provide a way of earning some interest in a relatively safe way.
The effect of tying up funds in time-deposit accounts is to take those funds out of circulation. That is a useful tool for the government to keep money out of circulation so that it can deficit spend on other programs and not be inflationary. It’s like the way the Bush-Cheney administration was able to deficit spend on the Iraq war while encouraging the buying of imports to counter inflation that otherwise might occur. It didn’t occur.
I needn’t have to remind you that with China and Japan owners of a large portion of these securities, they have acquired their dollars by selling us cars, microwaves, clothing, shoes, electronics. They don’t want to turn around and lessen their dollars’ value by flooding circulation with their dollars while the country is deficit spending on wars in the middle and far east.
So, if you buy up all the investors’ securities (for what purpose, I can’t imagine) you have to pay those investors their dollars, and what will they do with them if there are not safe securities to invest in?
As I say, the deficit spending securities have been mostly bought up by the Fed, which ultimately will swap them for new securities from the Treasury, so that the Fed can have sellable securities, especially during inflation.
The point is that deficit-spending debt is of minor importance and still well managed by Treasury and the Fed.
Why don’t we try to educate the public and our politicians about the folly of these arguments grounded in the belief that the national debt is about to bankrupt us as a nation.
That said, I agree that a $1T platinum coin might allow Treasury to fund without borrowing. Although it would have to be broken down into settlement balance dollars by the Fed and then transferred to the Treasury’s general account. Will the Fed cooperate? But to speak of this coin as the solution to reducing or eliminating the national debt seems irrelevant to being able to avoid the deficit ceiling since the national debt issue is really moot. You also would have the private banks’ objections to deal with, not just the Fed’s, because they want their free interest. That’s what they got in 1917 after years of lobbying to force Treasury to borrow instead of creating greenbacks.
Bonds are liquid BUT the liquidity is already in the economy. When they are sold money comes out of the economy and it does not get back in until maturity or when the Fed buys them. With Spending = taxes + tsys sales the Federal govt has zip power to alter the money supply. Fiscal policy can only move money around in the economy and does not alter the quantity. Period.
Once upon a time I believed in the platinum coin idea. But I came to see that it was unnecessary if not damaging to the economy.
Unnecessary because the govt and the Fed as an agent of the govt have no problem in either rolling over the debt (with securities swaps–Treasury) or buying back the securities with money created by the Fed out of thin air. More recently I have realized why the Fed’s buying is not inflationary. It is because it is paying off a private bank loan to the Tsy that the bank created out of thin air to get the money for the loan–and the loan is then cancelled between govt and bank, and the dollars used to pay the loan from the bank are extinguished after the securities are given to the Fed.
The Fed now has these securities. They may or may not be mature. But it will keep them until they mature and, *if there is inflation* it will take them to the Tsy to swap them for new securities. At this point the older mature securities are extinguished by the Tsy. The Fed takes the new securities and sells them to private investors to drain dollars out of circulation in the fight against inflation.
The platinum coin would be exchanged with the Fed for Federal Reserve dollars. Those dollars would be placed in a special account of the Treasury at the Fed.
Now, if the Tsy was so stupid as to use that money to buy back all the securities, it would be giving the investors back their dollars and what would they do with them? They would buy into our economy. Otherwise they would have no value to the investors. If done when there is already inflation, that could heat up the inflation even more. Securities are means for draining money out of the economy. I just didn’t think there was enough political support for the idea. Yes, it would be great if Treasury had again its powers to create green-back dollars without debt (beyond accepting taxes in them or paying back with other dollars their face value.
Stan, you say “…Now, if the Tsy was so stupid as to use that money to buy back all the securities, it would be giving the investors back their dollars and what would they do with them? They would buy into our economy.” Remember it is a 30-year process, not an instantaneous event. So it would be a draining over 30 years and only some nominal 500B$ per year assuming the Treasury did their job and sold them for a nominally consistent year by year payback. The congress should be able to accommodate that with relative ease with taxes and spending. Additionally, by paying off the debt, the Fed would be crippled and unable to manage the money supply via OMO. That means the congress would have to shoulder that responsibility to maintain price stability and employment. Additionally, that payback money did not come from folks who want o go shopping at Walmart. It was idle money and would probably end up in banks, CDs, commercial bonds, etc. I don’t think it would push up demand in any significant way.
“The effect of tying up funds in time-deposit accounts is to take those funds out of circulation. That is a useful tool for the government to keep money out of circulation so that it can deficit spend on other programs and not be inflationary.”
MMT disagrees for reasons I have already given.
I had sensed that disagreement in some sectors of MMT but did not know it was a “chiseled in stone” doctrine. The differences in views could be related to a very legitimate difference. All of these items are inherently stochastic and instantaneous values will differ from short term averaged values and will or can differ from long term averaged values. The MMT position can be very correct for instantaneous values and even for short term averaged values but I believe it cannot be accurate for long term averages.
I don’t recall what your reasoning on this was. Why would MMT object to tools at the Fed for fighting inflation by draining money out of circulation during inflations?
“Bonds are liquid BUT the liquidity is already in the economy. When they are sold money comes out of the economy and it does not get back in until maturity or when the Fed buys them. With Spending = taxes + tsys sales the Federal govt has zip power to alter the money supply. Fiscal policy can only move money around in the economy and does not alter the quantity. Period.”
This is true. When ownership of bonds is exchanged in the market, then the previous saver wish to increase liquidity to save in another form of asset, to invest, or to consume. This ability is not inhibited by the amount of Treasuries, since the Fed can always purchase existing tsys to increase liquidity as needed. This is one of the principal functions of a central bank.
The operative factor is flows of funds. Savers in tys can readily obtain liquidity for desired use of savings in tsys by selling them in the market. There are always buyers in a highly liquid market and the dealers also act as market makers. In the even that the system threatens not to clear, the Fed can step in and purchase tsys for its book adding more liquidity to the system.
“I had sensed that disagreement in some sectors of MMT but did not know it was a “chiseled in stone” doctrine. The differences in views could be related to a very legitimate difference. All of these items are inherently stochastic and instantaneous values will differ from short term averaged values and will or can differ from long term averaged values. The MMT position can be very correct for instantaneous values and even for short term averaged values but I believe it cannot be accurate for long term averages.”
The descriptive aspect of MMT that I have been discussing here is about operations actually used in government finance and banking under an endogenous money system with a central bank that imposes a reserve requirement using the US as an example. The US is a special case.
MMT also provides a general case description of central banking and banking under both fixed rate and floating rate systems, as well as descriptions of other special cases like the UK and EZ.
Can you comment on the time constants and stochastic aspects of the issue? I think it is generally accepted that the action to reaction time associated with open market operations is in the order of 6 months.
Sorry, I can’t. I am neither an economist nor in finance and am not up on the those kinds of studies. You’d have to ask in MMT economist.
My take on if from the MMT perspective is that conventional economists and people in finance mistake the role of the monetary base, taking it as causative (ex ante) rather than as a residual (ex post) resulting from the US reserve requirement. There is no monetary base in some countries like Canada since there are no required reserves. See Brian Romanchuk, whom I cited above.
The existence of a monetary base and its size really has nothing to do with banks’ ability to create deposits by lending or to get vault cash. The monetary base is just a relic of the gold standard that can be eliminated and some countries have done so. There is no “money multiplier,” for example.
Whether a country requires holding a percentage of settlement balances is irrelevant in that the system balances out at the end of every day. If there is a lack of liquidity, then the central bank just provides overnight overdrafts as needed.
Imposition of a reserve requirement just makes banking more expensive and banks pass this on to customers, so it acts like a slightly higher policy rate by increasing the spread.
Yes, the settlement balances just get shifted from Treasury spending into the economy into Treasury securities. In this way the amount of settlement balances remains the same.
This is something new. I’ve not seen anyone describe this before.
Correct me if I interpret this incorrectly: You are saying that after the banks note the increase in their own settlement balances at the Fed and match that in their local bank dollar settlement balances at the banks and credit the customers’ accounts in bank dollars, the Fed just takes those increased settlement balances it has on its books and uses them to buy securities from the Treasury. Actually it can’t be buying directly from the Treasury but via some of the primary dealers.
Why is that? Is it a more convenient way of storing the settlement balances than bank ledger entries?
Settlement balances are a Fed liability in deposit accounts at the Fed, where they are correspondingly assets of the owner of the deposit account. Bank deposits are bank liability and an asset of the owner of the account. Both the Fed and banks are banks, and banks create money “out of thin air” by crediting deposit account of clients.
The Fed cannot use the settlement balances in the banks’s deposit account at the Fed any more than the bank can use deposit accounts they create on their books. As assets of the client’s they belong to the clients.
The banks can exchange their settlement balances for cash to increase their vault cash for meeting window demand or tsys for their own balance sheet to gain more income from interest.
But the chief use of settlement balances is for settlement of transactions in the payments system, just as bank deposits are used by bank clients to settle accounts in the economy, for example, by using drafts against the deposits. Final settlement among banks of interbank drafts occurs in the payments system.
The Treasury also sells the tsys it issues by the Fed auctioning them in the payments system to the dealers, who pay for them using settlement balances they acquire in the payments system either through interbank transfers or borrowing.
When banks borrow from the Fed as lender of last resort, that is, through the discount window, the Fed creates an overdraft loan and credits the bank’s deposit account with settlement balances. The loan is a Fed asset and bank liability, and the settlement balances issued are a Fed liability and a bank asset. The bank is charged the penalty rate (discount rate) for the term of the loan. Loans in the payments system are overnight loans. The penalty rate is (normally) higher than the policy rate (aka “overnight rate”) to encourage interbank borrowing with the Fed always ensuring sufficient liquidity at the policy rate using OMO or POMO.
The money used in transactions in the economy is M!, that is, credits in banks’ deposit accounts that get drawn on and currency in circulation (cash). The cash comes from banks vault cash obtained by exchanging settlement balances at the Fed, and the deposits are created by banks crediting deposit accounts. Deposit account credits come from bank lending and borrowing, and government spending (including transfers).
Excepting cash for settling spot transactions, this is all occurs just by shifting entries in various accounts at banks, and using the Fed-run payments system for final settlement after netting.
This is the day to day operations of banks. The business if banks is actually risk management, which takes place through credit extension. When lending takes place, the lender and borrower assume risk of non-performance and the probability of that is reflected in the interest rate the lender requires to assume the risk. The real business of banking is assessing risk and banks that don’t get this right do poorly and may go under.
Oops. The M! in “The money used in transactions in the economy is M!” should instead be M1.
Tom, your reply to my query dated 19/08/2015 at 3:42 pm was quite informative, and I agree with it fully. However, it does not register with me as responding to my query. I was confused by your saying:
Yes, the settlement balances just get shifted from Treasury spending into the economy into Treasury securities.
I was hoping you would say that the banks would match the changes in the settlement balances spent into the economy via the banks, in their own bank settlement balances, which they would draw upon to put cash/bank dollars into customer accounts as directed by the Treasury’s checks to the customers.
In other words, I thought this was how to make sense of how settlement balances in the government sector get “transformed” into bank dollar settlement balances (not Fed settlement balances). Actually nothing is really transformed. The banks just have to note from the gains in their settlement balance accounts at the Fed, how much they need to add into their bank-dollar settlement balances (I presume there is such a thing). Then they would draw on these. (All this would be some private bank account from which funds could be drawn by the bank to credit some of its depositors’ accounts.) This account should match the reserve balance account of the bank at the Fed, but not contain federal settlement balance dollars.) Do you see what I am trying to do here to make sense of this?
I don’t see from the quote from you above how the Treasury securities have bearing on how federal settlement balance dollars ultimately from the Fed, go into the economy and (then?) into Treasury securities, since settlement balances can’t circulate, and securities can’t be bought with bank dollars. Only bank dollars circulate in the economy. But bank dollars are modulated by federal settlement balances at some level and in some portion of the economy. ?????
When the Treasury spends it directs the Fed to credit banks’ settlement balance accounts at the Fed with the amount of the invoice or transfer to the recipient.
The amount of settlement balances is a Fed liability and Treasury asset. The Fed is in effect transferring the Treasuries asset balance to the banks’ balance. This means that the bank receives an asset from the Fed that the bank then transfers to the recipient of the spending or transfer.
The bank does this by crediting the recipient’s deposit account at the bank. This deposit is a bank liability and a customer asset.
The bank has an asset in the form of the settlement balances and a liability in the form of a deposit entry.
This is how the funds get from the Treasury to the recipient.
The way the funds get to the Treasury account for spending it through the Treasury issuing securities that the Fed auctions to dealers. The dealers use Fed liabilities to purchase the securities that are Treasury liabilities and bank assets. The Fed then credits the Treasury account for spending.
In effect, the Treasury then has Fed liabilities as assets to spend which it received through the auction payment in settlement balances.
The amount of settlement balances spent by Treasury is the same as the amount of settlement balances that the banks used to purchase the tsys at the auction. So the amount of settlement balances in the monetary base remains the same in aggregate after the operations are completed as before. That is to say, the securities issuance drains the settlement balances added by deficit spending from the monetary base. The reason for this is to enable the Fed to adjust the monetary base to hit its target rate using less OMO/POMO than it would without the drain. So the drain serves as an interest-rate setting operation.
The increase in nongovernment net financial assets in aggregate, which is the fiscal addition from deficit spending, occurs through the Treasury issuing securities. As the economy needs liquidity, the Fed purchases these securities to increase the monetary base to meet the need. This just means that all transactions clear all the time seamlessly, while the Fed adjusts the size of the base not only to meet liquidity needs but also to hit its policy rate target.
The accounting shows the flow of funds through various accounts but that involves myriad transactions daily. This gets summarized as aggregates. Only the aggregates (stocks) and the change in aggregates (flows) get reported, not the individual transactions.
“I don’t see from the quote from you above how the Treasury securities have bearing on how federal settlement balance dollars ultimately from the Fed, go into the economy and (then?) into Treasury securities, since settlement balances can’t circulate, and securities can’t be bought with bank dollars. Only bank dollars circulate in the economy. But bank dollars are modulated by federal settlement balances at some level and in some portion of the economy. ?????”
‘Bank dollars” in deposit accounts are IOU’s of the bank to the customer that the bank promises to exchange for either cash at the window on demand or settlement balances in the payments system after netting.
This is the hierarchy of money. Bank dollars in deposit account have value for customers because customers trust the bank to exchange those bank dollars (entries) for cash (money token) or else for Fed liabilities in the payments system for final settlement after netting by clearing their drafts on deposit accounts.
The only money that is a “thing” is cash. The rest is just entries on various spreadsheets in the nongovernment banking system and at the central bank as the government’s bank.
The nongovernment banking system and the payments system operated by the central bank are two different systems that use the same unit of account and interface through the payments system.
There is a hierarchy in that both the banking system and government (Treasury and central bank) use the same unit of account (USD in US) but only the US government can issue its liabilities (settlement balances, cash and tsys). Therefore, while banks can create their own liabilities, they ave to obtain government liabilities that they don’t create in order to meet their promise to customers to exchange their liabilities as deposit account entries for government liabilities as cash at the window or as settlement balances to clear customer drafts on deposit account using the payments system after netting.
I think I am getting it. Please correct me where I am way off base below:
While the Fed can conduct the auction without the banks by dealing just with the primary dealers, taking advantage of their overdraft privileges at the Fed, I will describe what I think is more common, banks borrowing (using their overdraft privileges at the Fed) to obtain settlement balance dollars with which to buy securities for deficit spending. The banks want free interest.
When banks ‘lend’ the money to the Treasury for its deficit spending, the banks go to the discount window of the Fed and borrow the price for the securities that the banks are informed won their respective bids. There
should be no maturity date for replacing these loans. Banks will just pledge the securities as collateral should they default (which they technically will).
The banks get settlement balance (dollars) from the Fed as a loan.
The banks then go to the auction and pay the primary dealers the prices for the securities they won at the auction, using settlement balances from the Fed.
The dealers transfer the settlement balance dollars from the banks to the Treasury, which in turn gives the securities to the respective banks.
The Treasury then spends the settlement balance dollars it receives into the economy.
The Treasury does this by crediting the banks’ settlement balance accounts at the Fed. The banks see the additions to their settlement balance accounts at the Fed and match the settlement balances of the banks at the Fed with settlement balances in bank created dollars at the banks. Next the banks note from the checks it receives from the Treasury which customer accounts need crediting. It credits these customer accounts and debits the bank-dollar settlement balance accounts equivalently.
When customers want to spend the money from their accounts they seek cash from the bank. Customer fills out a check or withdrawal slip, and bank either transfers the money from the customer’s checking account to another customer of the bank or the check is sent to another bank where the recipient banks. … (I;m going to skip on these details to get to the next important point):
In the meantime the Treasury and the banks reach an accord to their mutual advantage, to roll-over perpetually the debts for the securities with swaps of new securities from the Treasury for mature securities from the banks, plus interest. The Treasury gets the principal it needed for deficit spending and the banks get the free interest at each roll-over. The principal will never be paid back to the banks because the roll-overs perpetually extend the maturity date farther and farther into the future on the securities. So, in effect, the money received for deficit spending becomes debt-free.
But the Fed can always demand the securities be given to it, and the securities themselves extinguish the debt to the Fed. Or the Fed might buy the securities with POMO. That will provide the banks with the money with which to pay off the loan from the Fed. The same thing will happen if the securities are bought by the Fed unconditionally with QE.
O.K., is this close to what you teach as doctrine for MMT?
“While the Fed can conduct the auction without the banks by dealing just with the primary dealers, taking advantage of their overdraft privileges at the Fed, I will describe what I think is more common, banks borrowing (using their overdraft privileges at the Fed) to obtain settlement balance dollars with which to buy securities for deficit spending. The banks want free interest.”
This is not the normal way it happens. The banks pay a higher rate for borrowing through the discount window in that the Fed normally sets the discount rate aka penalty rate higher than the policy rate that it sets as its target in the overnight interbank market. Banks generally borrow from each other in the interbank market or from the Fed using repo (tsys as collateral).
When the Treasury rolls over securities, the Fed ensures that there are sufficient settlement balances Treasury account by having the Treasury switch funds from a TT&L account to the TGA and supplementing by conducting auctions of the new issue. The Fed ensures that there are sufficient settlement balances in the interbank overnight market by purchasing existing tsys if needed, which increases the monetary base since settlement for the purchase is in settlement balances.
This is just cash flow management that banks and business do all the time to ensure that their payments clear on time.
All that has happened after the roll over is that the system has the same amount of tsys as previously, e.g., the same amount of T-bills. The Fed is rolling over a mountain of short term T-bills every day. The roll overs amount to trillions of dollars a year but no one even notices that the Fed is doing this by providing the liquidity needed for cash flow without lending directly to the Treasury.
If a bank happens to come up short on required reserves at the end of a period, then the Fed supplies an overdraft at the penalty rate.
“So, in effect, the money received for deficit spending becomes debt-free.”
It’s not “debt-free” to the government since Treasury continues to pay interest on the rolled over securities, just as it did on the maturing securities that get rolled over. In a roll over the maturing security is paid off and the newly issued security is sold to replace it. In aggregate, the amount of settlement balances that were injected into the monetary base by the redemption are immediately used to purchase the new issue. While this is just a swap, so to speak, it is not a direct swap since the process goes through the market and may involve different owners of the old and new issue.
The reason that parties hold securities rather than currency is that they prefer interest to liquidity. This is called liquidity preference and it is a changing variable.
“But the Fed can always demand the securities be given to it, and the securities themselves extinguish the debt to the Fed.”
I am not sure what you mean here. The Fed always purchases tsys with settlement balance when it acquires tsys for its own account, and when the Treasury purchases the tsys, they are not extinguished under the present rules other than in a conceptual consolidation of government books. That is to say, when the securities mature then the Treasury has to direct the Fed to debit is TGA. This credit to the Fed extinguishes the settlement balances as a Fed liability and the monetary base is not affected.
Otherwise you are close to the MMT view of banking operations.
I am not sure what you mean here. The Fed always purchases tsys with settlement balance[s] when it acquires tsys for its own account, and when the Treasury purchases the tsys, they are not extinguished under the present rules other than in a conceptual consolidation of government books. That is to say, when the securities mature then the Treasury has to direct the Fed to debit i[t?]s TGA. This credit to the Fed extinguishes the settlement balances as a Fed liability and the monetary base is not affected.
What I meant was based on assuming that the Fed lent the bank the price of the principal plus whatever penalty prior to OMO (sounds dumb that the bank would be penalized for borrowing to do the Treasury a favor of buying the security), with the security to be bought pledged as collateral in case of default (which is inevitable). The bank will not have the settlement balances to pay back the loan from the Fed, because the money (settlement balances) it uses to buy the security were spent away into the economy by Treasury. That would be why the bank would readily agree to the Treasury’s request to roll-over the security, because each would be unable to pay its debt obligation, and better to simply roll-over the security in perpetuity and leave the principal unpaid. Still the Fed would not lose in its loan because it could simply demand the security’s surrender to clear the loan to the bank. The Fed could then take the security and swap it for a new one with the Treasury (no settlement balances change hands) and hold on to the new security until inflation arose. The Treasury at that point would extinguish the mature security returned to it. (At the time inflation arose, the a mature security would be swapped for new securities again to allow the Fed to have a sellable security in that case).
If the Fed simply intervened and bought the security from the bank (in QE), that would give the bank the settlement balances with which to pay back the loan. But the effect in either case is the same. The Fed gets the security and the bank is relieved of the Fed’s loan to it for the security. A mature security is equal to the same amount of settlement balances in value and thus is a perfect exchange for settling the loan from the Fed.
That the deficit spending money spent by Treasury into the economy is debt-free refers to the fact that none of that principal spent is to be paid back because the roll-overs are intended to be done forever.
Ellen Brown and her followers are under the belief that all money created is debt, and believe it is a Ponzi scheme because the interest cannot be paid because not enough money is circulating to cover both the principal and the interest in aggregate.
So, in opposition to that view, it is important to point out that deficit spending becomes free of the debt on the principal. Interest, of course is paid, but the interest is new money borrowed independently by Treasury from banks and managed in the same way as the principal by roll-overs.
Deficit spending is also debt free when the Fed works with the primary dealers instead of banks using the dealers’ overdraft privileges with the Fed.
Dealer borrows from Fed settlement balances to pay for the Security. The Fed then buys the security from the dealer. Fed swaps the security for a new security with the Treasury, which effectively extinguishes the debt obligation to any holder of the initial mature security. The dealer has settlement balances now to pay off the loan via the overdraft and does so. The deficit spending money is now free of debt to anyone, except for the general obligation of all money to be accepted in payment of taxes, fees, and fines to the Federal government.
The banks pay a higher rate for borrowing through the discount window in that the Fed normally sets the discount rate aka penalty rate higher than the policy rate that it sets as its target in the overnight interbank market. Banks generally borrow from each other in the interbank market or from the Fed using repo (tsys as collateral).
I need clarification: Banks…borrow from the Fed using repo (tsys as collateral).
Is this is where a bank borrows from the Fed in OMO using the tsys
as collateral, so that if the bank defaults, the Fed demands surrender of the security, which itself will extinguish the loan. But is this just the ordinary loan plus penalty which is larger than the policy rate or is it at the policy rate.
When banks borrow (settlement balances) from each other do they independently set interest rates on the borrower to pay, or is there some regulated rate set by the Fed?
I’m trying to figure out when one method would be better than another in specific cases.
Tom, I said earlier
“The effect of tying up funds in time-deposit accounts is to take those funds out of circulation. That is a useful tool for the government to keep money out of circulation so that it can deficit spend on other programs and not be inflationary.”
There are several things up there to disagree with and numerous reasons might pertain, but I don’t see what it is. You need to explain yourself more.
I thought this was an MMT way of looking at the advantage to the government of having securities to sell to investors: When money is tied up in securities (savings), that is money taken out of circulation. There is a place for that.
My understanding of how inflation results is that it is when there is excess money in circulation (beyond the amount that would just permit full production and full employment at stable wages and prices). So, other things being equal if you returned all those dollars bound up in securities to the investors they could turn around and do something like (1) investing those dollars in our country or (2) buying goods manufactured from our country, or (3) seek to buy new securities from Treasury. Suppose we are already deficit spending and have raised the level of money in circulation as a result to the full production/full employment level, and all these dollars are returned to the investors and they pick the first two options (they are not necessarily mutually exclusive), wouldn’t that be inflationary?
I wrote a letter to Chair Janet Yellen at the Fed describing, as I understood it, the procedure by which primary dealers act as intermediaries with overdraft privileges at the Fed, in the purchase of securities from the Fed. Today I received the following letter which corrects some of my understanding:
BOARD OF GOVERNORS
FEDERAL RESERVE SYSTEM
Washington, D.C. 20551
DIVISION OF MONETARY AFFAIRS
Dr. Stanley Mulaik
191 Vistawood Lane
Marietta, GA 30066
Dear Dr. Mulaik:
Chair Yellen asked me to respond to your recent letter in which you raised questions related to federal deficits, the mechanics of Treasury debt auctions, and a set of ideas thathas been popularized as “Modern Monetary Theory.” On the question of the mechanics of Treasury debt auctions, there are a number of inaccuracies in the description of the process that you passed along based on the MMT website discussion. Here is how the process works.
The Federal Reserve acts as a “banker” for depository institutions, the federal government, and selected other financial institutions. Just as households and businesses maintain transaction accounts at commercial banks, these institutions maintaintransaction accounts at the Federal Reserve. These accounts are typically called “reserve accounts” and the balances held in these accounts at the Federal Reserve are commonly referred to as “reserve balances.”
Of note, primary dealers do not have reserve accounts at the Federal Reserve; all of the transactions of primary dealers are settled through correspondent banks. So contrary to some of the MMT website discussion you mentioned, primary dealers are never overdrawn at the Federal Reserve-they do not have reserve accounts with the Federal Reserve that can be overdrawn.
When the Treasury has expenditures that exceed receipts, it typically must borrow in the public debt markets to cover the shortfall. The Treasury issues new debt at public auctions on a regular basis for this purpose. The range of bidders at these auctions includes primary dealers, foreign governments, mutual funds, life insurance companies, pension funds and many other types of investors. Bidders at these auctions have arrangements in place to pay for the securities that they have been awarded at auction.
Typically, the “settlement” of a Treasury auction occurs a few days after the auction is held; at settlement, new Treasury securities are transferred to the securities accounts of the winning bidders. These securities accounts are maintained in the so-called “national book-entry system” that records the ownership of Treasury securities outstanding. In many cases, Treasury securities purchased by a final investor—-a mutual fund for example—are held in a book-entry securities account maintained through its correspondent bank. The Federal Reserve, acting as fiscal agent for the Treasury, processes the instructions that result in payments flowing to the U.S. Treasury associated with new debt issues. The Federal Reserve debits the reserve accounts of the correspondent banks of the investors that have purchased new securities, and there is a correspondent credit to the reserve account of the U.S. Treasury at the Federal Reserve. The correspondent banks that have received a debit to their reserve account at the Federal Reserve then pass a corresponding debit to the deposit accounts of their customers that have purchased new Treasury securities. Finally, the Treasury effectively pays down the temporarily elevated balance in its reserve account at the Federal Reserve to cover the excess of expenditures over receipts. At the end of this process, reserves held by the banking system are unchanged and the Treasury has issued more debt to the general public to finance the deficit.
It bears emphasizing that at no point in the Treasury auction process does the Federal Reserve temporarily purchase or sell Treasury securities to facilitate settlement on behalf of private investors nor does it provide credit temporarily to facilitate their purchases of Treasury securities.
I hope this information is useful. For additional technical information on the Treasury auction and settlement process, an excellent reference is the paper by Kenneth Garbade and Jeffrey Ingber at http://www.newyorkfed.org/research/current_issues/ci11-2.pdf.
James A. Clouse
Division of Monetary Affairs
Tom, do you have any comments to make about this?
Interesting. I’ll have to think about it and maybe ask some questions of those who know a lot more about operations than I do.
I haven’t answered in a couple of days because I am jammed right now.
Of interest to our discussion, Mr. Clouse confirms that Treasury auctions are settled using settlement balances in reserve accounts at the Fed that are Fed liabilities and bank assets rather than bank deposits that are banks’ liabilities. Financial institutions either must have the settlement balances in their accounts from normal operations or else borrow them in the interbank market.
As I said in previous comments, if an institution is short at the end of a period, then the Fed loans overnight by overdraft at the discount rate. This is rarely needed since bank have ALM departments to ensure that sufficient settlement balances are in the the banks’ reserve accounts for settlement, and the Fed ensures that there are sufficient settlement balances in the system to clear by adjusting the size of the monetary base using OMO and POMO, for example, repo and reverse repo.
You had asked a question on repo and reverse repo in a previous post that I did not have a chance to answer yet.
It is explained here.
As I understand you, your calculation show that there are not enough settlement balances to clear some auctions to make room for deficit spending. I said that this can only occur at the point of the debt limit, since Treasury can always emit public debt and the Fed will always ensure that settlement balances are available to purchase the issue at auctions by using OMO and POMO, and as the lender of last resort the Fed will always ensure that the system clears if a solvent institution with an account at the Fed comes up short at the close of period by providing overdraft at the discount rate. So there is never a situation that a Treasury issue will not clear for lack of settlement balances in reserve accounts at the Fed. The only constraint on the US government to issue US currency is the debt limit, which is not an operational constraint that is part of the system but rather a political restraint voluntarily imposed by Congress, which Congress has consistently changed to prevent a US default.
BTW, here is the Fed itself on PD’s
“Primary dealers serve as trading counterparties of the New York Fed in its implementation of monetary policy. This role includes the obligations to: (i) participate consistently in open market operations to carry out U.S. monetary policy pursuant to the direction of the Federal Open Market Committee (FOMC); and (ii) provide the New York Fed’s trading desk with market information and analysis helpful in the formulation and implementation of monetary policy. Primary dealers are also required to participate in all auctions of U.S. government debt and to make reasonable markets for the New York Fed when it transacts on behalf of its foreign official account-holders.”
I understand how busy you must be. Take your time.
One thing I note in the above letter from Mr. Clouse: He does not say much about how the Fed may interact with the correspondent banks. I’d like to pursue that further with him. We recognize how Fed and Treasury must be kept independent, and this means the Fed cannot buy securities directly from the Treasury. So, the Fed needs intermediaries. Are the Correspondent Banks the intermediaries in this case? They have reserve balance accounts at the Fed. And which are these banks? Will the Fed use repro and reverse repo with them?
I am a bit bothered by his seeming to treat those who buy securities for deficit spending as investors, like Chinese lending money to our government to help pay for deficit spending. Even with their having dollars, I doubt if we would sell them deficit spending securities. Maybe they wouldn’t want to buy them in the first place. Considering that an investor would not get his principal back if he ‘lent’ for deficit spending, because the securities would be perpetually rolled over, unless the Fed intervened and bought the securities outright with money it created out of thin air, I doubt if an investor would like that.
“One thing I note in the above letter from Mr. Clouse: He does not say much about how the Fed may interact with the correspondent banks.”
“Correspondent bank” just means a bank with a deposit account at the Fed that handles transactions for financial entities that don’t have deposit accounts at the Fed.
I find this somewhat strange. Citigroup Global Markets Inc. and Goldman, Sachs & Co. are listed as primary dealers, for example. Citi and GS are two of the largest banks in the US. It would certainly seem that they have reserves accounts at the Fed rather than going through correspondent banks.
So I don’t know what he means when he says, “Of note, primary dealers do not have reserve accounts at the Federal Reserve; all of the transactions of primary dealers are settled through correspondent banks. So contrary to some of the MMT website discussion you mentioned, primary dealers are never overdrawn at the Federal Reserve-they do not have reserve accounts with the Federal Reserve that can be overdrawn.”
At the time of the financial crisis Bernanke and Paulson called the heads of the big banks into tell them to use the discount window even through they didn’t need to as an example to other financial institutions. The discount rate was dropped to the FFR at the time eliminating the penalty, IIRC. The heads of Citigroup and GS were among those at the meeting.
The only way I can make sense of this is that the primary dealers associated with Citi and GS are subsidiaries that use the entity of which they are subsidiaries as their correspondent bank. That is a distinction without much of a difference.
Moreover, it doesn’t make any different whether the PD’s deal directly with the Fed or through correspondent banks. For example, the Treasury uses the Fed as it’s bank but as far as recipients of government spending and transfers are concerned the payment come from Treasury. The actual transactions take place in the payments system through the Fed and correspondent banks but the transfer is really among the actual parties. It doesn’t substantially affect anything.
“We recognize how Fed and Treasury must be kept independent, and this means the Fed cannot buy securities directly from the Treasury. So, the Fed needs intermediaries. Are the Correspondent Banks the intermediaries in this case?”
Not really. The dealers place the order and the correspondent bank executes it in the payments system with its settlement balances account and settles separately with the dealer.
Correspondent banks are generally the very large banks. That’s why I don’t’ understand why Citi and GS, which are PD’s, would need to go through a correspondent bank.
“They have reserve balance accounts at the Fed. And which are these banks? Will the Fed use repro and reverse repo with them?”
All institutions that have deposit accounts at the Fed can act as correspondents for others, I believe. The Fed uses repo and reverse repo with these institutions. For example, if a bank sees that it will come up short on its reserve requirement it can borrow in the money market overnight or from the Fed using repo. Finance is largely about managing short term liquidity. The Fed also uses repo and reverse repo for OMO.
“I am a bit bothered by his seeming to treat those who buy securities for deficit spending as investors, like Chinese lending money to our government to help pay for deficit spending. Even with their having dollars, I doubt if we would sell them deficit spending securities.”
All Treasury issue result fro deficit spending. The Treasury only issues securities to deficit spend.
“Maybe they wouldn’t want to buy them in the first place. Considering that an investor would not get his principal back if he ‘lent’ for deficit spending, because the securities would be perpetually rolled over, unless the Fed intervened and bought the securities outright with money it created out of thin air, I doubt if an investor would like that.”
Anyone holding a tsys can sell it in the highly liquid Treasury market instantaneously. No one is forced to roll over a security with it matures. It is a voluntary choice.
For example, when Chinese companies sell more in the US than they spend, then the Chinese firms have USD which they use to convert to yuan at the central bank of China (PBOC). The PBOC can choose to save in the USD they exchange for yuan, or sell in the forex market. If they choose to save, then if they keep the funds in a deposit account at the Fed, they earn no interest.
So they buy safe assets in which to save that pay interest by transferring the funds from their deposit account at the Fed to their securities account. Their preference for saving is in US tsys. The can roll them over when they mature by using the redemption funds to purchase other tsys to replace the maturing ones. They can convert the tsys to settlement balances in their deposit account by selling the tsys at market.
No one is ever locked into tsys. They are some of the most liquid assets in the world, as well as some of the safest. That’s why there is a big market for them. Large amounts are saved in US tsys to take advantage of the interest, even it is just overnight if money can be made by doing so. Even low interest on large amounts can be substantial.
“I am a bit bothered by his seeming to treat those who buy securities for deficit spending as investors, like Chinese lending money to our government to help pay for deficit spending. Even with their having dollars, I doubt if we would sell them deficit spending securities.”
This is not true of the case where the Fed swaps a mature security in its possession for a new security of the same face value from the Treasury. The Fed seems to initiate such a transaction, and it is not for deficit spending. It actually serves the objective of redeeming securities by returning them to the Treasury in return for new securities that may be sold by the Fed to investors and banks in OMO, POMO for controlling inflation.
Is there a law that says the Treasury may only issue securities to cover deficit spending?
Selling securities to investors does not constitute spending, but serves government objectives. The Treasury can only spend on what has been authorized by Congress. So, there are limits on Treasury spending; but I don’t see why there would be limits on Treasury lending or selling securities if these are backed by collateral of equal value, e.g. the securities themselves.
Why would the Treasury only issue for deficit spending? Treasuries are ways of sequestering dollars out of circulation in “time-deposit accounts” (or securities accounts) at the Fed. That is to the advantage of the government to avoid inflation when it has to deficit spend on important projects like infrastructure or defense. Selling securities serves a government objective other than deficit spending. The government’s intended deficit spending may not cause inflation because of a deflated economy. But if the economy is already having to absorb importers’ dollars, the government’s deficit spending may be inflationary when added into circulation. The importers may be enticed to buy securities instead of investing and buying in the economy and causing inflation that would reduce the value of their dollars. The investors apply for securities at the Treasury and the Treasury issues them to the investors.
Anyone holding a tsys can sell it in the highly liquid Treasury market instantaneously. No one is forced to roll over a security with [when?] it matures. It is a voluntary choice.
But if a bank has to borrow from other banks the reserve balances to buy the security, it may not have additional money with which to pay back the loan. So, the voluntary rational act would be to approve a roll-over. And Treasury would not have the revenues with which to pay off the debt and new debt as well so it would want to roll-over as well. But getting the interest may be the major incentive for buying the security.
BTW, what do you think of this proposition: when a bank borrows reserve balances from other banks to purchase a security, that does not take dollars out of circulation, because reserve balances other than currency aren’t in circulation ever, right? So that is not deflationary of reserve balances. Inflation only concerns dollars in circulation and levels of production and employment at stable prices and wages. And reserve balances are not withdrawn from circulation but cycled around within the Federal Reserve system. So, the aggregate pool of reserve balances in the banking system can be drawn upon forever by correspondent banks to fund new deficit spending, since the spending is converted by correspondent banks or commercial banks of government clients from reserve balances into bank dollars, and the reserve balances remain in the reserve system. Only the Fed can add as needed to the reserve balances of the system by creating reserve balances out of thin air.
Clouse’s letter leads me to believe that the correspondent banks exchange bank-created dollars for reserve settlement dollars. He essentially says that the Fed is not involved in helping banks and Treasury exchange dollars for securities in deficit spending. Another thing, it is said that reserve settlement dollars are used over and over and their quantity never changes. That means, to me, that when Fed buys securities from banks either in QE or OMO, their reserve dollars made out of thin air, are extinguished because they pay off a loan made by the bank to the Treasury. The bank must create the dollars as it usually does, out of thin air. Correspondent banks may help round up reserve settlement dollars from the banking system by borrowing them, then exchange them for the banks’ bank-created dollars. This is a kind of an immaculate conception for the Fed, in not having to create the dollars for deficit spending, but still able to cancel the government’s debt to the banks by buying securities with dollars the Fed creates out of thin air. But those dollars can’t go into circulation–reserve dollars never do that. They are extinguished as usual when a loan is repaid to the bank. So, what do you think? Do we have any alternative explanation with hard evidence to back it up? The Fed does not seem willing to easily publicize how it works, at least without a lot of obfuscation and smoke and mirrors.
Could it be that the reason correspondent banks are needed is to keep the Fed out of the picture. The reserve dollars pool at the federal funds level is finite and does not go into circulation, cannot cause inflation or be diminished or easily increased. (Fed would have to do it). So correspondent banks could simply call around among the US banking system for lending their reserve dollars to it for a client bank of the correspondent bank to be able to exchange its private bank dollars for reserve dollars. Every time the reserve dollars get exchanged for private bank dollars, the reserve dollars can be put back into the pool of reserve dollars to be accessed by all correspondent and big banks on an as need basis. Fed never has to put up money for a specific spending case. Its already available from the federal settlement reserve pool, which just cycles around all these dollars. The pool may be where all reserve-dollars go when extinguished. Sort of like hell for the Greeks.
Tom, why would someone wish to buy an already mature security? You would have to pay face value, and there would be no interest. Why would someone owning a mature security want to sell it for less than face value?
I suppose there is something I don’t know here.
Isn’t the answer obvious? A mature tsys has been “purchased” by the treasury and the face value plus interest earned in back in the economy in the account of the owner of record when it matured.
I don’t assume as maybe you do, Charles, that the banks will get to keep their principal on the loan along with the interest when the securities mature. That’s not the case with investor purchases of securities. Their dollars pre-existed the purchase of the securities. Bank dollars were created for the purchase of the securities. So, when the bank’s loan is repaid by the Fed at OMO public auction those dollars are extinguished, as should any monetary instrument created out of thin air when its debt obligation is satisfied.
Stan, as I see it the banks cannot loan money to the government that way for a very simple reason. The only way for a bank to create money for a loan is to do it via an account held at the bank and the government does not have accounts at commercial banks. It is a head scratching issue,,,I assume they purchase tsys with reserves thereby earning greater interest on their reserve holdings. By the way, I think it was 1991 when banks first held more govt paper than commercial paper. I remember an article in USA Today proclaiming the event.
charles3000 | 14/05/2016 at 2:28 am | said:
“Stan, as I see it the banks cannot loan money to the government that way for a very simple reason. The only way for a bank to create money for a loan is to do it via an account held at the bank and the government does not have accounts at commercial banks.”
Well, if what you say is true, it seems ordinary commercial banks cannot lend money to the Treasury, because the Treasury does not have an account at the commercial bank.
But maybe there is a way for the commercial bank to purchase the securities by using the services of a correspondent bank to exchange the commercial bank’s dollars it creates for the loan into reserve settlement dollars. The commercial bank’s dollars go into the correspondent bank’s vault cash and the reserve settlement dollars go into a reserve account at the commercial bank. The commercial bank can now draw from the reserve settlement account and write a check for the proper amount to the Treasury. Treasury will receive the check electronically and deposit the amount in its general account at the Fed. In return, the security is noted in the name of the commercial bank and the and stored in a spreadsheet for securities at the Fed. The Treasury will then complete its spending with dollars from its general reserve account at the Fed, which it will instruct the Fed to distribute to the reserve accounts of the commercial banks of recipients of the government spending, and from there exchanged into circulatible dollars deposited in the checking accounts of the recipients.
Does this sound plausible and feasible?
The tsys issues tsys to fund deficits. The accumulated tsys are the public debt that is also assets of nongovernment. When a tsys matures, the Treasury redeems it at face value. The interest was also paid when due. So both principal and interest are paid down.
The Treasury then has a choice. It can either reduce the public debt by not issuing a new tsys to replace the previous one, or else maintain the public debt at its present level by issuing a new security to replace the maturing one. In that case the funding used to redeem the maturing security that got added to the monetary base is just transferred to the purchase of the new security and the MB remains the same. The second alternative is called a roll-over.
If the Treasury decided to redeem a tsys and not roll it over, then it needs settlement balances to clear the transaction. If the Treasury is in deficit mode, it cannot pay down the public debt because it needs to fund the redemption. To pay down the public debt, the government fiscal balance needs to be in surplus so that the funding comes from tax receipts.
The government can be running a surplus and still decide not to pay down the public debt by continuing to roll over maturing tsys. That’s what happened with the Clinton surplus, for example, and then W “returned” the surplus to taxpayers through tax cuts that drew down the surplus.
Sorry, Charles, it’s not obvious to me. Maybe you can explain further.
I’m responding to Tom’s Anyone holding a tsys can sell it in the highly liquid Treasury market instantaneously. No one is forced to roll over a security with it matures. It is a voluntary choice.
Why would a second bank want to buy a security that had matured at a first bank for face value, which the first bank would want to get for it. There is no gain there in that transaction. Why give up an equal number of dollars for a security that is worth the same number of dollars on face value?
I think you are presuming the Treasury would simply redeem the mature security by paying the first bank face value. But Treasury is dealing with deficits because it has not collected enough tax revenues to cover the deficit spending from a previous budgetary cycle and a current one. And that is cumulative. So, how pay and with what money? It could issue new securities to borrow from banks the principal needed to pay the first bank and add in interest borrowed also from banks. But that is what a swap roll-over with the first bank would do between the Treasury and the first bank with the mature security. And the other bank is just first bank still. All that changes is just the maturity date to some more distant date in the future. Face values remain the same. Interest paid by Treasury to first bank comes from a borrowed-interest account also funded by issuing securities.
When tsys mature, they are automatically redeemed for face value. Not to do so would be default, as would not paying interest when due.
Tsys are traded before they mature and the interest goes along with the trade after discounting. Everyone gets the amount of interest as it becomes due as a matter of the trade. It gets figured in. The same security passes though many hands since they are used as a primary saving vehicle for large amounts. Corporations use T-bills instead of deposit accounts, for instance. Even though the interest is low, it’s better than nothing and adds up with large amounts. No CFO would ignore that.
China holds over a trillion USD in tsys as foreign exchange. The PBOC sold over a hundred billion some time ago and just sold another hundred billion recently. Yields have not risen as a result since demand for safe assets is high in the currently unstable global environment. Actually the ECB seems to have picked up the first 100+ billion. Central banks buy and sell foreign currencies to influence the fx rate of their own currency in order to promote trade by devaluing, for instance, or to stabilize the currency. US tsys are also used for foreign exchange and are sold when needed for international settlement in USD.
Stan, I believe you need to consider the effective mechanics and the over all magnitude of the operation. In a bygone era, before the Internet and digital communications, I am sure checks were printed and signed at the Treasury then mailed to the holder of the bond who could then get in his A-model Ford, go to a bank and get cash repaying him for his loan to the government at interest. But nowadays with the ubiquitous Internet and digital communications, I am very certain treasury bonds are repaid via a computer program. I am an amateur programmer and I could write such a program. A data base of bonds with maturity date, maturity value and holders bank account is all that is needed. On the due date the program increases the holders bank account the appropriate amount and the bond is paid. The program also would show the increase in treasury bonds for sale to effect the “roll over” operation. It would be foolish to not do it that way. Hence, bonds are “in maturity”, I would guess, for no more than milliseconds.
When tsys mature, they are automatically redeemed for face value. Not to do so would be default, as would not paying interest when due
Does the security not say, to effect, (in the earlier paper forms) To pay the holder on demand the face value of this security after maturity?
If so, then the holder can wait until it is desirable to demand payment.
Who automatically redeems them? Treasury or Fed? The Fed can create the money, but the Treasury can’t, and because it had to deficit spend, it likely doesn’t have the revenue to pay off the demand from the holder.
If the payer can’t pay, that is default, but the holder first has to present the security along with a demand for redemption. So it is up to the holder when to demand payment. But it would only be proper to demand after the maturity date.
Then the Treasury may refer the demand to the Fed.
I don’t believe that the Treasury is issuing paper securities any longer. They are held in securities accounts at the Fed, and they are redeemed when they come due, just as interest is paid as due.
Tsys are redeemed by the Treasury directing the Fed to debt the maturing tsy owner’s security account (through the owner’s bank’s account at the Fed) and credit the owner’s deposit account. This takes place through the bank marking up the customer’s account on the bank’s spreadsheet to reflect the transaction that takes place on the Fed’s spreadsheet in the payments system. The entries are as follows: Bank’s security account at the Fed debited and the bank’s deposit account at the Fed credited. The customer’s security account at the bank is debited and the customer’s deposit account at the bank is credited.
To do this the Fed must correspondingly debit the Treasury General Account. Treasury either gets the funds for this by issuing a new security (roll-over) or from taxes to pay down the pubic debt.
Treasury never has an issue with the redemption process or interest payments unless the debt limit prohibits issuance of more tsys and Congress doesn’t act.
In such a case, if the president chose to use the 14th amendment, he or she would direct the Treasury secretary to direct the chair of the Board of Governors of the Fed to directly credit the Treasury account.
Charles and Tom: By referring to the conditions under which the original paper securities were written: e.g. (only approximately from memory), “The United States government promises to pay the bearer of this security on demand the full face value of this security on maturity,” I was considering the promise, not how it would be implemented. Of course, we have computers, and they can automatize all kinds of things.
The holder has freedom to choose when after maturity to demand payment. See treasurydirect.gov.
But at TreasuryDirect the holder of the security has a choice: (1) roll-over the security (2) redeem the security and receive payment electronically.
I’m not sure what the Fed does with a mature security it buys (with money created out of thin air). I do know that it will (when?) exchange (swap) the mature security for a new one with new maturity date and thereby allow Treasury to extinguish the mature security that has come back into its possession.
I suppose the Fed can hold on to ‘immature’ securities until they mature, and do the roll-over. Or it can, before maturity, do a sale of the security in OMO and POMO.
I believe the Fed, under law, must transfer all interest received on tsys they hold less “expenses” to the US Treasury.
” I suppose the Fed can hold on to ‘immature’ securities until they mature, and do the roll-over. Or it can, before maturity, do a sale of the security in OMO and POMO.”
Fill in your details below or click an icon to log in:
You are commenting using your WordPress.com account. ( Log Out / Change )
You are commenting using your Twitter account. ( Log Out / Change )
You are commenting using your Facebook account. ( Log Out / Change )
You are commenting using your Google+ account. ( Log Out / Change )
Connecting to %s
Notify me of new comments via email.
Notify me of new posts via email.