Thursday, March 08, 2012

The Myth of High Powered Money

What is G-T?

When the Government spends (credits money in a non-Govt bank account) but doesn't tax it all bank (debiting money in that non-Govt bank account) what do you call the residual left in the non-Govt sector? On the Government side this is called the deficit (flow) or the national debt (stock). But what do you call it on the non-Government side?

I refer to it as "Net Financial Assets (equity)" because 1) we're talking about a financial asset here, not a real asset, and 2) the associated liability is out of sector. Therefore, unlike financial assets created in sector, the liability portion is booked as equity, not non-equity liability.

You cannot right call this savings (or net savings) because this causes all kinds of confusion as documented in the S=I debates rocketing around the internet these days. It does get close to the lay person's intuition about what "savings" are, or even what "nominal savings" are, but I don't think that term has been generally helpful.

You cannot really call it the nominal equity base that the private sector then leverages via the horizontal channel (gross financial assets) because all kinds of other things make up the equity value in non-Govt balance sheets.

You cannot call it "high powered money" or M0 because that actually refers to reserves, and reserves are a reactive function in that G-T generates a bank deposit, which generates a reserve by double entry book keeping. That reserve then either needs to be drained or it doesn't. But the causality is clear: NFA(e) creation causes reserve creation, but reserve creation does not necessarily lead to NFA(e) creation. NFA(e) is the private sector's nominal nest egg.

(Note: I'm ignoring exports etc. as I'm including foreign sectors as "non-Govt", because by Govt I mean the entity that is the currency monopolist).

Labels: , ,


Blogger Ralph Musgrave said...

Winter, I call it high powered money, and I don’t understand your reasons (beginning of your 2nd last paragraph) for NOT calling it HPM. Can you expand on your argument?

10:28 PM  
Blogger NeilW said...

You can't call it equity either. Equity is a liability in accounting, not an asset and will cause terrible confusion.

About the closest analogy I can think of is 'preference shares'.

'Bank Reserves' is the government sector account name on the government sector balance sheet (since in accounting reserves are generally an equity account).

On the private sector balance sheet they are really some sort of 'currency issuer preference share', whereas government bonds are really a 'convertible currency issuer preference share'

A lot of the confusion comes from using the name of the liability/equity account on the government sector side for the name of the asset account on the private sector side (and vice versa).

I think compounding that would make things worse.

11:53 PM  
Anonymous Anonymous said...

Ralph, net financial assets are not 'high powered money' because that term is already used to describe M0, which changes from day to day as the central bank trades financial assets while the net amount of financial assets held by the private sector stays the same. M0 is a gross measure of currency in circulation plus central bank reserves, NFA is a net measure of private financial assets minus liabilities.

Why would they be the same?

1:04 AM  
Anonymous Anonymous said...

I'm not sure its wrong to call it HPM or monetary base. Every check the government writes initiates a payment order process that results in the debiting of a Treasury account and the corresponding crediting of some bank's account. The bank starts by crediting its depositor's account but that in turn generates a payment order by the bank on the Treasury that is settled and cleared by the Fed. The only exception is that some of the checks are cashed for physical currency, and the bank might then replenish the physical currency by demanding its payment from the government in physical currency rather than a credit to its reserve account. But either way it is HPM, right?

Similarly, tax payments by check also initiate a payment sequence that results in a bank's reserve account being debited and the Treasury's account being credited. Or I suppose you can pay in physical cash. Either way its a drain in HPM or base money.

Reserve building is often seen as a reaction to the growth of deposits because one way deposits grow is through the growth in aggregate commercial bank lending, and that aggregate growth in lending necessitates a subsequent growth in reserves, enabled by the Fed, so that banks can meet their reserve requirements and make all of their payments smoothly within the constraint of the Fed's target interest rates. The expanded lending might not expand reserves, if the banks are starting from a position of excess aggregate reserves. But when it does boost reserves, the causation is from the deposit boost to the reserve boost, so the reserve boost is a reaction.

And as we have seen (Edward Harrison wrote about it again yesterday), boosts in reserves don't have a causal influence on boosts in deposits. Base money or HPM can grow dramatically without a corresponding increase in broad money, i.e. deposits.

But when deposits grow as a result of net Treasury spending - spending not offset by equivalent amounts of taxes - then the growth of HPM is much more automatic and direct, not reactive. The net spending is in that case a straightforward injection of HPM. It's not like banks autonomously expanding their liabilities through lending, and then seeking reserves if required to meet their reserve requirements and boost their capacity to make payments. Instead, when they accept that government check into the deposit account of their customer, the government at that point simply owes them money with no further discretion or causal gaps.

5:17 AM  
Anonymous Anonymous said...


The deficit doesn't have anything to do with HPM, unless that deficit is funded by issuing new HPM, and, as rule, deficits are not.

6:27 AM  
Anonymous Anonymous said...

And whether a Treasury deficit corresponds to a net government injection of monetary base to the private sector requires looking at the consolidated balance sheet of the whole government, not just the Treasury. Current operational rules require the Treasury to plug the hole in its budget with borrowing from the private sector. So, in the current period, if the Fed doesn't increase its purchases of treasuries, then whatever portion of Treasury spending is not offset by current tax revenues will be offset by a drain of monetary from private sector purchases of securities.

However, if the Fed then purchases those securities, the monetary base loss from securities purchases is replenished, and the total government position represents a net injection of monetary base or HPM.

As a result the Fed incurs an additional "liability" on its own balance sheet, since currency in circulation and reserves are booked that way by the Fed's accounting conventions. But it is a pretty silly kind of so-called liability, because the existence of currency in circulation doesn't correspond to something that the Fed still owes the private sector. The Fed is effectively an infinite pool of endless dollars, and when it spends on an asset purchase or to make an interest payment, it just dips into the infinite money pool and pulls some out. So the view of the Fed as being in a state of positive or negative "equity" is an antiquated accounting convention that does not really correspond to any kind of economic reality.

6:40 AM  
Anonymous Anonymous said...

The deficit doesn't have anything to do with HPM, unless that deficit is funded by issuing new HPM, and, as rule, deficits are not.

The portion of additional public borrowing that is purchased by the Fed represents a net injection of HPM. When that purchase is made, the Fed credits private sector accounts, and part of the Treasury liability to the private sector is converted into an intragovernmental liability.

6:47 AM  
Blogger paul meli said...

From where I sit…

NFA dollars are a subset of NFA.

NFA dollars are a subset of HPM.

NFA bonds are not a subset of HPM.

Credit money does not expand NFA.

In the interest of clarity there seems to be a need to differentiate between NFA, HPM and credit money, because they are not the same.


6:51 AM  
Anonymous Anonymous said...


The level of treasury debt and the level of base money are two distinct values.

When the government runs a deficit, the private sector does not accumulate base money, unless that deficit is funded by issuing base money. Since, as MMTers should be aware, the government does not fund its deficits by issuing base money.... Well, I'm sure you can do the maths.

The supply of base money is independent of treasury borrowing. It is determined by the Fed's reaction function and private demand for notes and coin, for the most part.

Fed operations are just portfolio adjustments from the point of view of the private sector. No one has more or less "NFA" as a result of a Fed repo.

6:54 AM  
Blogger paul meli said...

"…the government does not fund its deficits…"

True. This would be FAPP impossible.

Deficits = creating new NFA out of thin air.

NFA dollars can be increased without selling treasuries to the public.

7:16 AM  
Anonymous Anonymous said...

vimothy, I believe the Fed's stance is not independent of treasury spending. Suppose next year, the US Congress, for whatever reason, decided to quadruple the size of the US government deficit and eliminate all debt ceiling restrictions while not increasing tax revenues, so that the quantity of Treasury debt issuance quadrupled. How do you think the Fed respond? Would it increase its volume of purchases of treasury securities or wouldn't it?

If it does immediately increase its volume of purchases then the US government as a whole is making a net injection into the monetary base.

It could decide not to do this, in which case the added injections of currency and reserves caused by the expanded spending are offset entirely by removals of currency and reserves to purchase securities that mature in a later period. And if the Fed were overcome by price stability vapors, they might even offset the larger deficit by selling securities to drain reserves. But if they are in stance where the full employment mandate is regarded as more important than the price stability concerns, then they will adopt a policy that allows the monetary base to net increase to grow aggregate demand.

7:23 AM  
Anonymous Anonymous said...


Your counterfactual begs the question.

If the Fed continued to target interest rates and allow the public to hold whatever stock of notes and coins it desired, then it wouldn't do anything different to what it does now.

If the Fed decided instead to monetize the debt, then it would do so.

7:40 AM  
Blogger paul meli said...


Does it matter if the Fed's stance is or is not independent of Treasury spending?

Just trying to understand your line of thinking here ignoring the fact that Fed actions can be counter to the goals of Treasury spending.

In what ways do you see the Fed having an effect on the Treasury other than by undermining spending policy goals with monetary policy?

7:45 AM  
Anonymous Anonymous said...


And do you think that what the Fed decides to do is independent of the fiscal stance of the Treasury? What the Treasury is doing is a huge part of the overall macroeconomic environment, and so responding to the Treasury position falls within the scope of the Fed's dual mandate.

If the Fed does not accommodate a fiscal expansion with a monetary expansion, then it is basically forcing the government position into one of Ricardian equivalence, with the Treasury deficit just driving a a recycling on money rather than an expansion of money. If it does this during a period of high unemployment and low inflation or deflation, it is violating its mandate.

8:12 AM  
Anonymous Anonymous said...

Paulie, one thing that became very clear during the debt ceiling debate is that to understand the MMT model of the government as a currency issuer or monopoly producer of the currency, with the power to increase the private sector's supply of net financial assets, it is necessary to look at the entire consolidated governmental sector, with the Treasury and Fed combined.

The fact that the government sector as a whole is a currency issuer, and not just a currency user, does not mean that each part of the government is a currency issuer. The Treasury is required to function under US law as something very much like a currency user. Just like a private sector currency user, if its spending exceeds its revenues, then it either has to draw down its savings or take on debt. The Treasury is not permitted overdrafts on its account at the Fed and not permitted to borrow directly from the Fed. And in practice, coin seignorage plays a minimal role. So in effect a Treasury deficit triggers an overdraft on the private sector, i.e. more debt to the private sector.

But the government sector as a whole is not like a household at all. It can run a "pure deficit" It's total receipts from every source can fall short of its total spending. But for this to happen requires the cooperation of the Fed. The Fed performs various operations that inject and extract currency and reserves from the private sector, and whether the government sector as a whole is net supplying additional currency and reserves, net extracting currency and reserves or holding the quantity constant depends of what the Fed does.

8:31 AM  
Blogger paul meli said...


Ok, we're on the same page.

The only question I have is would the Fed ever not accomodate Treasury needs and/or has the Fed ever not accomodated the Treasury in the past?

In other words is there or is there not a defacto consolidation or does the fact that there is a slight chance, however unlikely, that the Fed would not accomodate Treasury negate any model assuming consolidation under current law?

8:56 AM  
Anonymous Anonymous said...


We're discussing the stock of HPM specifically, not the role of the Fed in general.

Your second para is wrong. I'm not sure where you're getting it from, but it's wrong.

Nothing the government or the Fed can do can ever force the the economy to behave according to Ricardian equivalence--none of which has anything to do with base money.

9:04 AM  
Blogger paul meli said...


I'm with Mr. speak on the thrust of his post. He's speaking my language.

One caveat - I believe I've worked through a chain of logic in previous discussions demonstrating (to my satisfaction anyway, FWIW) that savings (and investment) in the traditional economics context and in the MMT/NFA context are the same in substance and in magnitude.

I'll leave it to Neil Wilson to reconcile this with contradictory evidence (apparently) gleaned from NIPA and FoF data by others since he seems to be the most knowledgable source in this area.

9:43 AM  
Anonymous Anonymous said...


I'm not "getting" what I'm saying from anywhere. I'm just giving my own opinions, although obviously I have learned most of what I know from reading other people.

My understanding is that Ricardian equivalence thesis claims that (i) government spending is entirely funded by transfers from the private sector - either immediately through taxation, or on a delayed basis through borrowing and subsequent taxation to pay the debt - and so most of that spending is offset by a corresponding short-term or long-term reductions in private sector spending, and (ii) people tend to know this is the way things work, and form their expectations accordingly. They thus expect that any increases in the relative rate of government spending will be followed by increases in the relative rate of taxation. The private sector thus offsets the effects of the increased government spending by increasing their own saving rate in preparation for the higher taxes.

Now I agree that there are several things wrong with this account - particularly some of the assumptions about expectations built into (ii). But the main thing wrong with it, from an MMT point of view, is part (i). As a currency issuer rather than a currency user, increases in the relative rate of government spending are not always offset by transfers from the private sector to the public sector - either by immediate tax increases in the short term or by borrowing in the short term and tax increases in the long term. Some of that increase in government spending is due to a net injection of additional money into the economy.

However, if the government chose to do so, it could voluntarily impose an ordinary budgetary constraint on itself: force itself to act just like any old firm or household. In this imagined world, the central bank never buys debt from the government, directly or indirectly. It only supplies additional reserves to the private sector through the banking channel in response to growth in private sector demand for credit. And that's it. Otherwise the domestic private sector and the government are just a closed loop of interacting currency users.

If we lived in such a world, it would not be very far from a Ricardian world.

Fortunately, by my lights, we do not live in such a world. Instead the government partially "finances" expansions of its deficit by part of the government issuing debt to another part of the government, and then by rolling these bookkeeping maneuvers over indefinitely. The intra-governmental debt-issuance is mediated by private sector middle-men, who are permitted a small cut for their services. But essentially, and over the long term, the operation consists in the government spending more, through all channels, than it takes in, through all channels.

9:47 AM  
Anonymous Anonymous said...

The only question I have is would the Fed ever not accomodate Treasury needs and/or has the Fed ever not accomodated the Treasury in the past?

I really don't know that much about it Paulie. But it seems to me that if the Fed governors were convinced that the Congress and Treasury were out of control inflationists and budgetary incompetents, they might choose to offset expanded federal deficits by declining to increase their own purchases of government debt, thus guaranteeing that everything injected into the private sector on the one end was drained out on the other end. They might even drain out more than the Treasury is injecting in, to offset inflationary effects from a shift from idle savings into active spending.

Remember that Paul Volcker is often credited with intentionally driving the economy into recession to combat inflationary effects - even though the government was running deficits.

Bernanke, on the other hand, has been very cooperative by buying up massive amounts of government debt in two quantitative easing programs, which among other things resulted in the erasure of a significant part of the Treasury's interest liability to the public.

9:59 AM  
Blogger winterspeak said...


the point of my post was that what is called "high powered money" is *not* the same as NFA(e). You and I are in agreement, they have nothing to do with each other.

Госбанк thought they were the same, and they are not. NFA(e) is what Mosler calls "vertical money".


Equity is a liability in accounting, and yes this will confuse the layman, but I'm taking things up a notch and expecting basic understanding of double entry book keeping.

To the layman you can call it "savings" and they will have the right sense for it. But to a more technical crowd "savings" causes confusion.

ALL: Yes, the Ed Harrison post is timely. HPM may not be that HP.

10:15 AM  
Blogger NeilW said...

" But to a more technical crowd "savings" causes confusion."

If it's a choice between 'savings' and 'equity', I'll take equity.

In fact I'd rather use 'wibble' than savings...

11:01 AM  
Blogger Госбанк said...


Госбанк thought they were the same, and they are not. NFA(e) is what Mosler calls "vertical money".

Just to clarify.

Госбанк did not think they were the same.

What I meant, and I am sorry that I was not clear enough, was that the moment the treasury deficit spends, the NFA flow is materialized as base money flow into the aggregated private sector account from the aggregated treasury account. Assuming for simplicity no taxation and only borrowing from the private sector (bond sales), one can imagine this circular flow of base money from the bond buyer's bank via the treasury account and into the recipient's bank. In this idealized picture, very small amount of base money is needed to achieve a substantial cumulative NFA flow because base money inflows and outflows to/from the treasury cancel each other (I am aware of TT&L accounts, but their role is precisely to smooth out base money flow imperfections).

Of course, the only apparatus capable of changing the stock of base money in the banking system is the Feds as Vimothy has already indicated.

Now, having thought a bit more, I realize that stylized picture above may not be very good at all, and that base money existence is an unimportant banking system implementation detail to make gov. generated claims injection into the private sector possible, along with other minor services such as currency/notes provision and monetary policy as well as being the LLR conduit.

The recipient of gov. spending, it is interesting to note, will most likely have "horizontal money" in his account represented as his bank liability resulting from the treasury crediting the recipient's bank with an equal amount of base money.

The net result of government spending would be person A holding a bond, person B holding horizontal money that person A used to hold assuming all the proceeds from the bond sale went to B. The amount of base money in the "system" is unchanged and its role is completely hidden from the observer.

12:17 PM  
Blogger Ralph Musgrave said...

Winter says, “You cannot call (G-T) "high powered money" or M0 because that actually refers to reserves..” (I assume Winter is saying that HPM/MO is the same as reserves). I disagree.

HPM/MO is not the same as reserves because physical cash in the hands of the public is HPM/MO, but it is obviously not bank reserves.

HPM/MO according to Investopedia is “any liquid or cash assets held within a central bank and the amount of physical currency circulating in the economy”. A definition I agree with (except that they’ve missed out physical cash held by commercial banks – or perhaps the latter is included in what they call “circulating in the economy”).

So $X worth of G-T will certainly add $X to HPM/MO, but it could in theory just add $X to cash in circulation with the public and add nothing to bank reserves. (In practice $X worth of G-T would doubtless add a bit to both cash in circulation AND add a bit to bank reserves).

That is all supported by an economics text book called “University Economics” (by A.A.Alchain) which says that monetary base equals vault cash held by commercial banks plus commercial bank deposits at the Fed plus currency in hands of public.

2:50 PM  
Blogger winterspeak said...

Ralph: It is true that M0 includes more than reserves, but it does include reserves. And NFA(e) might create reserves as a consequence, but it really isn't about reserves at all.

Also, when people take cash out of the bank, then that money is unplugged from the reserve system and is "off the grid". When you withdraw cash, you reduce a liability at the bank, and the bank reduces a reserve asset, and then Fed reduces a liability on their side, and there's no reserve asset credit anywhere in the system until the cash gets redeposited somewhere.

When the Mint prints currency, it's better thought of as supermarket inventory.


NFA(e) is a stock, not a flow, and is equal to the national debt. The deficit is the delta NFA(e) -- so deficits increase it, surpluses decrease it.

Note that all govt bond sales do is change the term structure of outstanding NFA(e) to the extent that they were purchased by non-leveraged buyers. If the bond sales were purchased with leverage (as you correctly suggest is possible), then NFA(e) remains both constant in size and unchanged in term structure.

And yes, as you and vimothy stay, strictly under current official institutional arrangements, only the Fed can carry a negative NFA(e) position. The Treasury cannot, although there's an open question of whether, if push came to shove, Bernanke would bounce Tim's cheque.

The Treasury, under current institutional arrangements, operates just like a bank in that it needs to sell bonds. The Fed is truly the currency issuer.

3:25 PM  
Blogger Steve Roth said...

Steve Waldman has talked about the notion of govt-created money as constituting equity in government -- similar in that a dollar gives you a claim on govt assets and the products of those assets, but nothing like real ownership rights. You can't ask for your brick from the Pentagon. Also there's not fixed date for retirement of dollars or stock shares.

I understand the discomfort with the "equity" usage because there's the "equity" that a shareholder "owns" as an asset on the left side of their balance sheet, and there's the "shareholders' equity" liability on the RHS of the firms' balance sheet. (Or governments', as suggested here.)

So it can be confusing. I also would like to find a better, more immediately comprehensible term for that RHS liability.

>The Treasury, under current institutional arrangements, operates just like a bank in that it needs to sell bonds. The Fed is truly the currency issuer.

This, I think is why the thing is so bloody confusing. The treasury (a sovereign issuer) has the constraints of a bank, while the fed (a bank) has the freedom of a sovereign issuer. Sheesh.

5:05 PM  
Blogger winterspeak said...


Steve Waldman's "equity" concept may appeal to his norms, but I think a more dispassionate approach shows it is not correct.

For example, if a Corporation sets up some plantation, and pays its workers in scrip that they can only use in the company store, would you say that those scrip wielding workers are in some way "equity holders" in the Corporation itself?

That position would seem surprising to both the workers themselves, who may feel they are being exploited by said Corporation, and the Corporations actual equity holders, who are profiting from the enterprise at large.

9:20 AM  
Blogger Steve Roth said...


I don't understand your analogy. Is the government here the corporation issuing scrip?

10:31 AM  
Blogger Steve Roth said...

"pays its workers in scrip that they can only use in the company store"

?? Dollars -- whether issued by govt or banks, doesn't matter once they're in the system -- can be used pretty much anywhere in the world.

"the Corporations actual equity holders"

Who are they, in this analogy?

11:59 AM  
Blogger winterspeak said...


I'm not talking about some abstract thing. These actually existed, and probably still exists.

A Corporation, such as a logging company, opens up a worksite somewhere remote in the woods. It pays its workers in a mix of currency and scrip. (It issues the scrip itself). The scrip is redeemable for goods at the local company store, but nowhere else. Since the work site is in the middle of nowhere, the workers don't really have any other way to buy stuff.

This corporation is a private company, and it has shareholders like any other private company.

In such situations, the corporation is creating and issuing its own currency (script). The workers have to accept this currency because they have no other choice. The scrip can only be used in company stories.

It is obviously ridiculous to think of this scrip as somehow being "equity" in the corporation.

The corporation needs to have issued more scrip than its collected back in order for the workers to have some stash of scrip on hand to buy stuff when they want. They would call this "savings" but, I'm calling it NFA(e).

2:06 PM  
Blogger Steve Roth said...

I get all that, but I don't think your analogy is illuminating. Because, there are no shareholders in government. SRW's thinking is that the public holding government scrip, in aggregate, constitute something ery like those shareholders -- notably including fairly limited voting rights over the "corporation"'s activities.

9:57 AM  
Blogger winterspeak said...

Yeah, but SRW's thinking is ridiculous as clearly, holding scrip doesn't give you the right to anything other than extinguishing an obligation demarcated in that scrip.

SRW may CLAIM that holding scrip SHOULD give you all kinds of rights that he finds politically beneficial, but that's very different from what actually is.

2:48 PM  
Blogger NeilW said...

Interestingly Ralph has put up a quote from Frederick Soddy - who wrote "The Role of Money".

' On p.40, under the heading “Money a Claim to What Does not Exist.” he says, “The essential feature of money is . . . that it is a legal claim to wealth over and above the wealth in existence, all of which in an individualistic society is already in the ownership of others independently of this claim.”'

That is what people believe, (and leads to the 'confidence' arguments), when 'de jure' all fiat money actually is is a Tax Credit.

1:26 AM  
Blogger winterspeak said...

Neil: Right. They cannot fathom just how much it really is just numbers on an excel spreadsheet.

It is the monopoly on violence which comes from (defines) Sovereignty which gives it value, and even then, only to the extent the Sovereign decides.

SRW is very uncomfortable with accepting concentrated Sovereignty.

9:27 AM  
Blogger ChrisJCook said...

Interestingly enough the undated credit created and issued by sovereigns against value received, and returnable against taxation was for 500 years known as 'stock'.

The phrase 'rate of return' does not refer to an interest rate but the rate at which stock - which was part of a split 'tally stick' - could be returned to the issuer who would match it against the counter-stock or 'foil.

Any economic entity can issue 'stock', and the 'stock' in a joint stock company was not the same as 'shares'.

Gilts - aka 'gilt-edged' stock are in fact not a debt claim but a form of equity, akin to interest bearing preference shares issued by governments.

While the 'scrip' referred to above as being issued by corporates is in fact 'stock' in its original meaning.

A misconception has crept in re tally sticks, which were an early accounting tool, because they could be issued either as a receipt (a memorandum tally) or record with no obligation, or as an IOU returnable against value and issued at a discount to value received (without a discount no-one would accept them).

2:17 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home