Jan 29, 2023
Welcome to episode 140 of Activist #MMT. Today I talk with Scott Fullwiler on his 2008 paper, Modern Central Bank Operations: The General Principles. Today's part one of a three-part conversation. Today in part one we discuss some generic but related topics, and then principles one and two. Next time in part two we cover principles three to six, and then in part three, principles seven to ten. My full and detailed question and summary list can be found at the bottom of these show notes (look below!). Also, be sure to check out the list of audio chapters to find precisely where each principle, and otherwise, can be found.
(Here are links to parts two and three. A list of the audio chapters in this episode can be found right below the resources section in this post.)
Today's principles one and two.
Principle one is that reserves can only be used for two purposes: Settling payments between banks, and meeting reserve requirements. (There's actually a third purpose, which is it's the only thing that can ultimately settle tax obligations to the state.) Knowing these are its only possible uses, when you hear, for example, that more reserves somehow increase a bank's liquidity, and that this in turn encourages banks to lend more to customers, which then in turn increases economic activity in general… you know they're wrong. The same is true with the reverse: that less reserves somehow discourages lending and reduces economic activity.
Principal two says that, because the central bank is the only entity capable of creating and deleting reserves, it has "a fundamental, legal obligation to promote the smooth functioning of the national payment system." Without a functioning payment system, society would, without exaggeration, break down. If a bank can't settle its payments with another bank, then everyone expecting a payment won't receive it, and everyone expecting payment from them also won't receive it. And on and on.
Trillions of dollars go through the federal reserve system every day. More goes through this system in the United States each week then an entire year's worth of GDP. Not to mention, the US payment system is central to most of the payments for the entire world, and so the US payment system breaking down would have global implications.
(As a brief side note, this latter point is leveraged by the United States to surveil and manipulate most nations around the globe. One example is how, when Iraq threaten to eject all US troops, the US responded by threatening to forbid Iraq from using its payment system, thereby potentially disconnecting it from the entire world. This is the big story that lurks behind the so-called petrodollar. Here is a fascinating video on this by the Wall Street Journal.)
And now, onto my conversation with Scott Fullwiler. Enjoy.
I have some questions before we get into the ten principals:
Pre-1: First, I'd like to start with a general question mostly unrelated to your paper: A common online theory is that the central bank doesn't answer to the government. Rather, the government answers to the central bank – and according to some, even directly to commercial banks. This means the government must borrow (in the personal sense!) from the CB or banks, which means the national debt and deficit, and bond vigilantes, are indeed a big deal. This also completely undermines MMT.
We're going to get into lots of details, but in general, how would you respond to that person? (Assuming they really want to know better.) Is there any instance in history where, when it really came down to it, the central bank didn't do what Congress or Parliament demanded of it?
Having a stable society requires a stable payment system, which, under our current institutional set up, only the central bank can do. Is it possible to have a stable society/payment system, and a dollar worth the same on both sides of the country, if the government had to answer to the central bank in that way?
Pre-2: Your paper, written in 2008, is called Modern Central Bank Operations: The General Principles. Can you tell the backstory of how the paper came to be, as you briefly mentioned in email?
Pre-3: As I understand it, horizontalists and structuralists agree that loans create deposits, but disagree on the how, where, and dwhy the reserves are obtained afterwards. Can you summarize the differences and the debate between the two camps, and also relate it to the chartalist view?
Pre-4: How do you know what you know? You interviewed CB employees? Looked at their balance sheets? Just logically it must be true?
Pre-5: It's been fourteen years and two major crises since you wrote your paper. How well do the ten principles stand up? If you wrote the paper again today, would there be any major changes?
THE PRINCIPLES
I'm going to summarize the ten principles in your paper as best I can, and describe some of their implications. Then I'll ask you to correct and elaborate as necessary. I'll also use some of the principles as an excuse to ask a question.
PRINCIPLE ONE
Reserves only serve two purposes: settling payments and meeting reserve requirements. Regarding the latter, there could be an arbitrary requirement that, for example, a bank must always hold an amount of reserves equal to 10% of the amount it has in deposits (perhaps immediately, or with a lag). In the absence of reserve requirements, the amount of deposits held by a bank is only very distantly related to the amount of reserves banks need to make settlement. This is because a newly created deposit for a newly created loan (or from new government spending):
So again, the existence or creation of new deposits is only very indirectly related to the need for more reserves.
A minor follow up: Banks require reserves to transact with entities other than itself. These other entities include other banks, and the government at all levels. What other institutions/entities require reserves for settlement? Foreign banks and governments?
PRINCIPLE TWO
As the only institution capable of creating and deleting reserves, the central bank has "a fundamental, legal obligation to promote the smooth functioning of the national payment system." As you say in the paper, "a nation's payment system is at the core of the infrastructure of the modern business world." According to the Federal Reserve's Board of Governors in 1990: "A reliable payments system is crucial to the economic growth and stability of the nation. The smooth functioning of markets for virtually every good and service is dependent upon the smooth functioning of banking in the financial markets, which in turn is dependent upon the integrity of the nation's payment system."
The amount of transactions settled each day is enormous. In the US in 2005 it was $2.1 trillion. Today I believe it's closer to $5 trillion. So, a sixth of the annual GDP of the United States, is processed each day by the central bank. Further, this is only a portion of the nation's transactions, because more are directly settled between banks through side agreements and internal systems.
The central bank is the only institution that can create reserves, and so, if we are to have a functioning society, it will provide the reserves needed by the banks, because it's the only thing that can settle those transactions. If a bank abuses these privileges (such as, they keep demanding more and more, because they keep committing crimes) then they could be shut down.
An analogy is how parents are the only ones capable of providing their children with food. Ultimately, it's provided based on the needs of the children. Parents will provide enough food in order for their children to remain healthy and not dead (and so they don't have to go to jail). It also implies a power struggle, such as when the children whine about being hungry, not out of actual need but as a form of manipulation. Of course, unlike the banks and their central bank, in most normal families, the children haven't paid off their parents. Also unlike banks, a child can't be shut down if they consistently misbehave – unless the parent really wants to go to jail and lose all their children!
PRINCIPLE THREE
Before I summarize this principle, can you talk about how the money multiplier view and fractional reserve banking are two sides of the same thing?
The principle:
The money multiplier not only doesn't limit bank lending, it's impossible for the central bank to directly target reserve levels, or the monetary base, at all. It's only possible to directly target the price of that money – the interest rate. The monetary aggregate can only be indirectly targeted, which is inherently unreliable. Even if the central bank could magically manage the levels of reserves, since banks are not reserve constrained, it wouldn't have any direct effect on bank lending anyway.
It's impossible for the central bank to control the level of reserves because there are many factors out of its direct control. This includes:
As we're about to discuss in principle four, all these activities must be continually offset. Attempting to target specific reserve levels can only serve to degrade its ability to manage these offsets, and so its target rate, and ultimately, the payment system.
PRINCIPLE FOUR
As in the previous question, the central bank does many things unrelated to interest rate targeting, and many other things happen out in the world that aren't directly in its control. This results in reserve levels moving in an unpredictable fashion, all of which must be offset if the target rate is to be maintained.
One of the things out of the central bank's control is government spending. The way the government spends occurs is mind twisting, and understanding it is key to understanding national accounting specifically and modern money in general.
The government itself has a checking account at its central bank, which in the United States is called the Treasury's general account, or TGA. This is the account where a number is raised in response to new spending voted on via the passage of a new law. [CORRECTION: As (needlessly!) required by law, the TGA is not raised except after tax and bond revenue is received.] When that money is distributed to someone in the real economy, that same number is lowered once again. This is a very nature of government spending.
Here's another example of this mind twisting: When the government sells a bond, it's paid for by the government. The government does this by withdrawing $1000 from its account, the TGA, and handing it to the central bank. So, to pay the bank – it's bank – it withdraws $1000 from that bank and hands it right back to the bank! Further, at some future date, the bank must then pay its profit to its shareholders, which is the government. How do they do this? By putting that money right back into that same government account! (Of course, no money is actually passed around, it's just a number going down there and going up here.)
(Also, the government's account can go deeply negative without much real-world consequence, but since negative numbers stress uninformed people out, we cater to (and leverage) that ignorance by making sure it stays positive.)
PRINCIPLE FIVE
Reserve requirements are related to interest rate targets, not control of monetary aggregates.
In one sense, what's having the purpose of having rules at all when it's guaranteed that the rule maker will do whatever it takes to ensure the rule followers always follow the rules?
It seems reserve requirements are a tool to buffer against sudden volatility, in the same way that TT&L accounts (as stated on page 607 in Stephanie Kelton's 2000 paper, Do Taxes and Bonds Finance Government Spending?) are used to buffer against volatility from government spending and redemption. These things don't stop the need for offsetting these activities (as in principle four), but it does make it possible to not have to do it at such quick, extreme, and unpredictable levels. In other words, these buffers don't change what the the central bank needs to do but it helps them see it coming.
I'm going to ask a mostly unrelated question:
Interest rates are for managing the target rate, which is for managing the stability of the payment system, which is for maintaining the stability of the entire nation.
Yet, at the same time, the CB is also mandated to manage (some definition of!) inflation, and the only way it knows how to do this is by adjusting interest rates. How can these tasks not conflict? If it's critical to keep interest rates stable (near the target, ideally zero from our MMT points of view), then during the Volcker shock, how could you possibly keep interest rates stable at such a high level? In that situation, it seems that banks simply settling their payments each day would be so expensive, they would have to pass much of that cost onto their customers through higher interest rates.
Raising interest rates:
PRINCIPLE SIX
Volatility in the target rate is only possible between the discount window's penalty rate at a maximum and the interest rate paid on reserves at a minimum. The way you say it in your paper is, "Potential volatility is determined by the width of the corridor."
Here's a question about the target rate and its corridor or band (with thanks to Andrew Chirgwin):
Let's assume a corridor with a width of .5%. So the minimum, the interest on reserves (IOR), is 1.75%. The target rate is 2%, and the penalty/discount rate is 2.25%. So, they're all different values.
If a bank is in need of reserves, it first turns to another bank. It may be a bank it needs to settle with, but maybe not. It may try to get all the reserves from one bank, or maybe a little from several.
In order to turn a profit, the banks with excess will make an interest-rate offer to the bank-in-need. That rate will be somewhere within the band. It won't be higher than the penalty rate, because the bank-in-need could just turn to the central bank's discount window and pay less interest. It won't be lower than IOR, because no bank would deliberately choose to lose money (that is, make less from the bank-in-need, than they would from interest paid directly on their reserves).
Within this narrow band, banks with excess may compete with one another in an attempt to get the business of the bank-in-need. So, although a bank may offer an interest rate of, say, 2.24%, which is just under the penalty rate, another could easily steal their business by offering 2.20%. The central bank is okay with this competition, because they know the interest rates will remain within the band.
What I don't understand is, the CB defends that band so that it remains within the minimum and maximum. So, why is there a precise target at all – and consequently, what's the point of potentially setting it equal to IOR? Clearly I'm missing something, because it's stated at several points in your paper that setting the target rate equal to IOR does make an important difference.
How does the central bank defend the precise target rate?
A somewhat related thought experiment, which may just be absurd:
What would some of the major consequences be if the discount window/penalty rate was set below IOR? (With the target rate between the two.)
PRINCIPLE SEVEN
In the context of monetary policy, the concept of "liquidity effect" is that extra reserves in the interbank market pushes down interest rates, which then stimulates banks to make more loans, which in turn increases economic activity. In other words, it's the false view that the interest rate is not something the central bank can arbitrarily decide, but rather something it can only control or defend by offsetting the effects of "market forces". Luckily, since the central bank is the largest currency user, it at least has a decent chance of success. (I know that's not what they mean but it's not far off!)
Specifically, the "liquidity effect" is the false belief that the only way for the central bank to "choose", or defend, its target rate, is to inject a potentially vast amount of reserves into the banks' balances. This will encourage banks to increase lending, which in turn will increase economic activity. This is called "easing". (QE is just a ridiculous amount of easing.) Removing a large amount, called "tightening", will discourage lending and economic activity.
In reality, the target rate is an arbitrary decision (a "policy variable") of the voting members of the central bank. The consolidated government has the infinite capacity to create and delete its own money and to sell and purchase its own bonds. This means it can effectively choose an interest rate for any bond at any maturity.
The false "liquidity effect" view also asserts the mere existence of more reserves in a bank's account makes banks suddenly need them; makes them want to use them. It strongly suggests that reserves can be directly lent to customers, or can be used for some purpose beyond settlement (and meeting reserve requirements). If my bank dramatically increased my personal checking account, then sure, that would indeed cause me to pay off my mortgage and probably hire some contractors to do fixes and upgrades to my house that at the moment, we can only dream about. But that's only because, for average people, deposits can be used for almost any purpose. [CORRECTION: Me getting money in my bank account, outside a loan, is net financial asset – a grant. The back being reserved is always an even swap. That's totally different.]
Beyond reserve requirements, the only possible use of bank reserves is to settle transactions – transactions that happened at some point in the past. It means the mere existence of more reserves has no direct influence on a bank's behavior. In other words, settlement – and therefore the amount of reserves needed – is endogenous. A bank's demand for reserves is vertical. It's decided on not by the government but by actual people choosing to take out a loan and a bank choosing to give them one
A final point: The false idea of the "liquidity effect", that the mere existence of new reserves incentivizes banks to issue more loans, evokes the concept of Say's law. Say's law is the false idea that supply causes demand, as if a new product appearing on a store shelf magically and magnetically attracts a new customer – who didn't even know the product was existed – to want to go to that store and want to purchase that product. (As if consumers are unthinking puppets and businesses their puppeteers!)
In reality, demand causes supply. In reality, loans create deposits. Those deposits will at some point likely result in some transactions with another bank, which the bank will need to settle. If they don't have enough in reserves, only then will they request more.
PRINCIPLE EIGHT
The quantity of reserve balances in circulation is primarily determined by the central bank's method of interest rate management.
The only uses for reserves are to settle payments and meet reserve requirements. If there are no reserve requirements, then there's clearly less reasons to hold them. As a simple example, if the central bank chooses to penalize overdrafts severely at the end of each day, then banks will demand much more reserves in order to buffer against that possibility.
If there were no reserve requirements, and both IOR and the penalty rate (and the target) were set to zero, then it seems there would be little to no uncertainty for banks. It would be free to purchase reserves from the discount window whenever needed. This seems close to, if not exactly, MMT's ZIRP.
If all three were equal but set *above* zero, then banks would make a profit on their reserves, and when in need of more reserves for settlement (again assuming no reserves requirements), they would pay that same rate at the discount window. (There would be little need for banks to lend to each other, because they could do no better.) So, again, it seems there would be little concerns from banks to make settlement or fear overdrafts. The only difference is the perpetual risk-free, effort-free interest income!
These are different methods the central bank can choose to manage the interest rate. What are some other important scenarios/methods and their practical differences, both from the banks and the central bank's points of view?
PRINCIPLE NINE
Under current operating procedures, the central bank's balance sheet expands and contracts endogenously while these changes neither create nor destroy net financial assets for the non-government sector.
In your paper, you say: "neither reserve balances nor the monetary base can be expanded or contracted exogenously by the central bank as long as the central bank's target rate is above the rate paid on reserve balances."
With our previous questions as background, can you elaborate on this?
PRINCIPLE TEN
This principle is basically distinguishing between the currency issuer and users
Central banks interest rate "matters" because banks use reserve balances to settle payments.
Banks and "market forces" do not control the interest rate. This is for the simple fact that banks must settle their transactions at the end of each day, those transactions can only be settled with reserves, and those reserves can only be supplied (created and deleted) by the central bank.
Also:
The banks don't control the central bank and its interest rate any more than average people control the commercial banks at which they have a deposit. Even the most powerful currency user has no power over the currency issuer, because their power largely comes from that issuer! (They were issued a lot, while the rest were issued less.) Any power the user has over the issuer is only because the issuer chooses for it to be that way.
FINAL QUESTIONS
If you could have your dream government, what economic and financial appointments would you make? What position would you want?
If those people got appointed, then what are some of the big changes we would see, particularly regarding monetary policy?