Between Depression and Hyperinflation
Contra Megan, there is a middle ground between a Depression and Hyperinflation:
Too much printing can generate hyperinflation, but too little printing leads to low aggregate demand, unemployment, and lost real resources -- the very thing a country should be looking to maximize.
But it is not true that loads of debt is just fine as long as you're borrowing in your own currency, except in the trivial sense that a government which borrows in its own currency can always resort to hyperinflation. This is rather like saying, "Don't worry about that cancer--you can always shoot yourself!" If you take too much advantage of the benefits of borrowing in your own currency, pretty soon you have trouble borrowing in your own currency, which means that practically, the distinction is not necessarily as strong as some people pretend.This is the conventional wisdom, believed by Austrian, Paul Krugman, and Greg Mankiw alike. A country that runs a fiat currency doesn't "borrow", it prints and unprints money whenever it spends and taxes. If it issues bonds, it is to change the term structure of extant pre-printed currency, not to print more, or "sterilize" outstanding money.
Too much printing can generate hyperinflation, but too little printing leads to low aggregate demand, unemployment, and lost real resources -- the very thing a country should be looking to maximize.
posted by winterspeak at
9:11 AM
111 Comments:
It all boils down to not understanding that private banks can put some of the money supply 'on ice'.
Once you realise that loans and deposits pop into circulation at the same time, you realise they can both pop out of circulation at the same time too.
The key mistake is to believe that when a loan is repaid the circulating money just moves somewhere else in the system. It doesn't - it turns into potential money stuck in a bank awaiting a willing borrower at the current price of money.
Between Printing and Unprinting, Borrowing and taxing, isn't Default an option? In my view, the following:
1. Investing in Sovereign debt is practically no different than investing in stocks, i.e. based on the greater fool's expectation = I will be able to get off this security before it bites the dust (in a default or the company slumps). When the debt of a Sovereign Gov has reached unbearable levels, default is an objective option, the creditor can blame themselves for betting on a risky horse when it was too close to exhaustion. Sovereign Bonds, especially of Govs that already have debt levels reached several years of tax collection, are as much speculation as buying stocks is.
2. 'Corporate borrowing costs will sky rocket if the nation in which they are incorporated sufferes a Sovereign rating downgrade' is taken as an axiom, but I don't think it is true always. Today's Corporate America is not dependent on Gov for raising (or guaranteeing) it's debt Capital, has a war chest of a few trillion dollars in cash-at-hand - so does not even need to go to markets for raising more debt for many years to come.
Neil: Not quite. When a loan is paid back, gross financial assets shrink. The money disappears as liabilities and assets are both marked down.
The "potential money" exists only in the sense that the bank's capital is now less levered, meaning it has more ability to lever by making additional loans and re-expanding its balance sheet.
OHM: Don't agree with 1., but I might agree with 2. If you buy the debt of a currency sovereign you are taking on inflation risk and political default risk (as Russia demonstrated in 98) but not operational default risk they way you are if you buy corporate debt or non-sovereign currency debt.
So, a currency issuer may CHOOSE not to pay, and you may not be able to buy much with what they pay IN, but the option not to pay is always a choice, never a necessity.
Megan McArdle? She's a complete ignorant. She doesn't even understand mainstream economics. I remember her complaining that the government can't even securitize future tax cash flows? Which begs the question why are US treasuries called securities?
The money only disappears because that's the 'accounting policy' you're sticking to.
Enhance the model with an 'inventory of credit' and the loans just become inventory when paid back, and inventory increases when the bank gets extra capital. The size of the inventory then explains the bank's 'hunger' to make new loans.
There used to be a time when purchased goodwill was written off against the balance sheet. It 'disappeared'. Yet with a change of accounting policy that is now held on the face of the balance sheet.
You can do the same with the 'credit licence' intangible asset a bank holds.
Stephan: I wouldn't be so hard on Megan. She took the same (wrong) macro classes I did. If you're in paradigm, the entire concept of "securitizing future cash flows" don't make sense at the Federal level.
Neil: The "accounting policy" I'm sticking to is how the accounting is actually done.
Banks expand and contract their balance sheets, they do no re-assign existing assets and liabilities. When a bank buys computers, it is a balance sheet expansion.
This is very different from a household.
But you can use whatever "accounting policy" you prefer.
Winter, RE: "but the option not to pay is always a choice, never a necessity."
Sometimes it can be a necessity - to avoid hyperinflation and collapse with monetization, given the size of the debt burden...and inability to raise taxes without seeing more than proportionate shrink in the economic activity base. So, When one invests in the debt of a Sovereign, one is placing the bet that the Sovereign will not reach such a necessity point nor will default as a choice....that's my 1. :-\
Ohm: If hyperinflation is a real risk, it means that either the population has waaay too much money, or real output is collapsing.
If the population has waaay to much money, then the sovereign, if it is sovereign, should be able to tax sufficiently to control inflation.
If the population has real output collapsing, then money supply (however you choose to define it) can stay static and you will still get hyperinflation.
But altering the term structure of outstanding sovereign liabilities (which is all that happens when a sovereign "pays back" its debt) should have no impact on anything. How could it? Why should it? It's just moving money from a savings account to a checking account.
But yes, Ohm, when you buy sovereign debt it is a bet that the sovereign will not dissolve. What does a Rhodesian Dollar get you nowadays?
"when a sovereign "pays back" its debt) should have no impact on anything."
[Assuming that only aggregate money is what matters and not it's distribution at the hands of the population]
It will not have any impact when the Sovereign pays it back (instead of spending it all, that it did in the past years) from it's current year tax collection. If it has to print new money to pay back (while maintaining same tax and spend), or pays from unspent tax collection accumulated in the past few years, it is increasing the aggregate floating money stock at the hands of the people.
Sovereign Default does not require a dissolution of the Sovereign, it can also occur and the Sovereign lives on. In fact, the United States recently went through a soft default through the debt limit drama, it's Debt rating got downgraded, it not only survived as a borrower in the market, but it's interest rates for fresh borrowings fell! Guess, the new interest rates, post default, depends upon the various options that money has for investment, and the post-default (perceived) health of the borrower relative to other seekers of Capital.
Since Sovereigns today are not on the Gold Standard, the necessity for defaulting is technically lower...as printing can be resorted to. But depending upon the enormity of the debt and (gloominess of the) growth outlook for economic activities, printing option to repay might have reached the point of being a bigger risk, than defaulting and starting fresh on borrowing (or running a balanced budget for a while, which becomes easier with lower or no interest payment as default has already been performed, or running small deficits for a while financed with printing). In the Gold Standard days, I believe, European Kings defaulted several times on their debt, and continued on without dissolving. It (default without dissolution of the currency area) just hasn't happened until now in a Western country on pure fiat currency, but might be quite pull-off_able.
See:
http://en.wikipedia.org/wiki/Sovereign_default#Consequences
and sections thereafter.
esp: "The citizen might feel the impact indirectly through high unemployment and the decrease of state services and benefits. However, a monetarily sovereign state can take steps to minimize negative consequences, rebalance the economy and foster social/economic progress (eg. Plano Real)."
Neil,
Winterspeak has it right on the accounting for capital and credit. In fact, the nominal amount of future credit that can be made available from a given inventory of excess capital is a function of risk assessment at the time the credit is approved, so it can't be determined ahead of time. The real inventory function is capital, not credit.
You can do it your way as well, but its not very close to actual bank accounting for credit, risk, and capital. Loans and money are definitely destroyed when a loan is repaid. Looks like you been influenced by Steve Keen, who's invented his own accounting along those lines.
"Banks expand and contract their balance sheets, they do no re-assign existing assets and liabilities."
You've missed the point that how the accounting is done depends upon the audience. Accounts are adjusted for their purpose - which is why financial accounts differ from cost accounts from tax accounts in their presentation yet are consistent with each other.
Somewhere in a bank somebody calculates the figure based on the capital ratios and provides it to someone.
So saying 'that is how the accounting is done' is not how accounting is done. And it certainly isn't how modelling is done.
Both views are correct and consistent with each other. So it depends on the audience and what you are trying to do with the accounts.
The ability to switch viewpoints is what makes MMT and Post Keynesian useful. Rigid thinking is turns great ideas into religious statements.
"In fact, the nominal amount of future credit that can be made available from a given inventory of excess capital is a function of risk assessment at the time the credit is approved, so it can't be determined ahead of time. The real inventory function is capital, not credit."
It doesn't need to be created ahead of time. But it does need to be calculated *at the time*. You need that number to be able to determine whether you can advance a loan and whether you can charge interest since that is, initially at least, a loan.
So it boils down to whether you want the figure explicit or implicit on the model balance sheet. Both are correct because the credit inventory number is derived dynamically from the capital position and the amount of outstanding loans.
And the only reason you have to worry about capital positions is because the credit licence (which is the real intangible asset IMV) links loan capacity to capital. If the credit licence was, say, a fixed amount then capital would not be the inventory item.
So another way of looking at it is to see the 'asset' as a valuation of the current credit licence with the revaluation written to a reserve (much as you would a property).
(And for the record what I'm doing here is exploring the unifying view that allows Keen's models to be consistent with the MMT horizontal model.
As an open-source advocate I'm looking to integrate Keen's 'fork'. It seems to me that people get are getting excited over mere viewpoint differences rather than anything substantive. I reckon one can easily be derived from the other - once you fix the basic double entry errors).
winterspeak: But altering the term structure of outstanding sovereign liabilities (which is all that happens when a sovereign "pays back" its debt) should have no impact on anything.
Just to clarify: Do you implicitly assume here that the holders of the corresponding assets are insensitive to the interest rate?
At the very least, it would seem that changing the term structure might cause some of the holders of the debt to go out and buy real things instead of holding on to the debt, if the interest rate decreases. So then the question would be one of how strong that effect is, right?
Neil,
"So another way of looking at it is to see the 'asset' as a valuation of the current credit licence"
OK, but excess capital is a single nominal number, while there is no corresponding single nominal credit reserve, because nominal credit availability depends on level of risk taken per asset.
But I agree generally, subject to that qualification.
JKH,
Thanks. That has helped enormously in determining the approach to take here.
winterspeak,
Thanks for hosting and thanks for humouring me while I evolve my thinking.
Ohm: A sovereign never "has" or does "not have" money. Every act of spending is money printing. Every act of taxation is money un-printing (in the fiat world).
The difference between money printed and money un-printed in a period is what the populace gets to keep in their wallet. Federal Deficit = Private Savings.
Inflation isn't caused by money printing. It's caused by too much money chasing too few goods.
Nicolai: holders of debt may seek yield (expansionary), but they also get less income, and so may spend less or try to save more (contractionary). Net net I think it's a wash.
Winter, I'd like to *qualify* parts of what what you wrote. That is,
A sovereign never "has" or does "not have" money *(Agreed)*. Every act of spending *not covered with taxes or borrowing* is money printing. Every act of taxation *if not then spent by the Sovereign* is money un-printing (in the fiat world).
The difference between money printed and money un-printed in a period is what the populace gets to keep in their wallet. Federal Deficit = *increase in combination of Private Savings & Private Consumption*.
Inflation isn't caused by money printing. It's caused by too much money chasing too few goods.....*which can be caused by too much money printed out that it did not convert to increased production of goods and services. In an extreme case, the currency might simply lose the faith and respect of the people, and collapse (rendered useless).*
*Ohm: Lastly, default on past debt too is a mechanism of money unprinting, like increased taxation/Gov spend (=paying down past deficit). none of the money unprinting mechanisms are needed to be considered when there is a good outlook for growth in output. The increased economic base yields improving deficit/GDP scenario, higher output counters hyperinflation that incremental printing can cause.*
Hi Ohm:
No, what I'm asserting is that *all* acts of spending is money printing, regardless of how much is in the kitty, or whether there is a bond issuance, or anything else. *All* acts of taxation, are equally, un-printing. The sovereign never stock piles, the sovereign never "borrows".
The Federal Deficit is the increase in private savings. Not consumption. Consumption can increase or decrease regardless of the deficit.
Inflation is caused by too much money CHASING too few goods, which can be brought on my too much printing OR too much chasing (increase in V) OR too few goods OR some combination of these. In your extreme case, the Sovereign can enforce taxes and make people respect the currency again.
Yes, default on debt is the same as money unprinting. But, money unprinting via taxes is likely deflationary, money unprinting via default is more likely to be inflationary IMO as it will increase V as people substitute out of the currency into other things.
@Ohm -- you started with:
"1. Investing in Sovereign debt is practically no different than investing in stocks, i.e. based on the greater fool's expectation = I will be able to get off this security before it bites the dust (in a default or the company slumps). When the debt of a Sovereign Gov has reached unbearable levels, default is an objective option, the creditor can blame themselves for betting on a risky horse when it was too close to exhaustion. Sovereign Bonds, especially of Govs that already have debt levels reached several years of tax collection, are as much speculation as buying stocks is."
From an MMT point of view, that's .. not true, as winterspeak explains.
Another way to express the MMT view is this:
Yes, the amount of non-government Net Financial Assets (NFA) matters. (Can also be thought of as Net "Default Free" assets).
Net Financial Assets is Government Bonds and Base Money (Bank reserves and cash) held by non-govt, less whatever has been lent by the Fed. (I think.)
While the amount of NFA matters, its composition is not that big a deal.
You can think of Government Bonds and Base Money as substitutes for each other. Both can be thought of as equally "default-free". Both are government IOU:s.
Does the non-government hold interest bearing bonds or zero interest base money? It's not that big a deal.
True, it matters for interest rates. And interest rates may matter a bit (although the effects are somewhat ambiguous).
But it's not like the Government bonds is a huge burden for future generations while currency is not. Both are financial assets for non-government, and neither is a burden.
Inflation? Well, if NFA increases, that may cause increased spending. That increases aggregate demand. And eventually, demand could surpass the capacity of the real economy to meet it. Inflation would ensue.
But still, whether NFA increases in the form of bonds or currency doesn't matter much (barring some ambiguous effects from interest rates).
Winter, RE: "Consumption can increase or decrease regardless of the deficit."
Sure, it can, so can savings (as saving-consumption-income-debt are mutually balancing faces of the same cube, or diamond, if you will).
"The Federal Deficit is the increase in private savings" is not a true statement, through the Bush years, there was increased deficit but not a commensurate increase in private savings, though there was a consumption boom with this deficit and wanton credit. It cannot be said that federal deficit will necessarily increase private savings, and private savings only, though sometimes it can.
----
Using your terminology,
1. All Government Spend is printing
2. All Government taxation is unprinting
1 & 2 together is redistribution
3. Default on debt servicing is unprinting
4. Government borrowing is... unprinting(?)
5. Servicing the debt is printing
So, it is the net incremental printing/unprinting that changes the active money stock that can have the effect of inflation (if little supply of real goods gets added) and/or growth (increase in production and private consumption) and/or change in private savings. The same devices above can also be used to rebalance the distribution of the money stock which has it's own effect on economic activities. It's not possible to simply equate or approximate federal deficit = increase in private savings, although there can be periods of time when it holds.
"A country that runs a fiat currency doesn't "borrow", it prints and unprints money whenever it spends and taxes."
Not if the central bank is prevented from crediting the government's account at the central bank, which is how modern fiat systems tend to be run.
"Not if the central bank is prevented from crediting the government's account at the central bank, which is how modern fiat systems tend to be run."
That's a sub-sector obfuscation device.
Consolidate the government and central bank balance sheets to the minimum set of accounts and you'll see the 'print/unprint' relationship happen - regardless of the sub-sector rules.
In the MMT model the currency issuer is part of the government sector and the effects tend to refer to the sector level balance sheet.
I am in awe of such flexible pithiness from Mosler, having reversed his position on the use of the previously banned “printing”, so long as it’s now combined with “unprinting”.
Deep stuff indeed. Full steam ahead.
It seems to me that there are two different pictures one can use to describe the role of money in consolidated government financial operations. But if I'm not deceived, both pictures are consistent with the MMT view of the world, and the choice of which one to use is more a matter of taste and explanatory convenience, rather than a substantive theoretical difference.
On one picture, government taxes and borrowings destroy or extinguish net money, and government payments create net money. Net money creation occurs during some period when more money is created than is destroyed. Taxes differ from borrowing in that taxes destroy net financial assets while borrowing increases net financial assets, i.e., in borrowing the government extinguishes some money from a lenders account, but replaces that money with a financial asset that is a promise of future money creation on the lender's account.
On this picture, the government never has money itself - as Neil says. Money is something that only exists in the non-governmental sector. Government monetary operations control how much money there is in the non-governmental sector, and where it is. But they do not have the result that the government either acquires or relinquishes money.
On an alternative picture, the government can possess money in its account. Taxing consists in a transfer of already-existing money from a non-government account into the government account, and payments consist in a transfer of already-existing money from a government account to a non-government account. But the government differs from other entities in having the power, at any time, to create or destroy money in its own account. On this picture, we think of monetary operations as creating or destroying money inside the government account, and then propagating the effect those changes to the non-government sectors.
Am I wrong in thinking that which picture one uses is just a matter of convenience?
This comment has been removed by the author.
This comment has been removed by the author.
No, and I wrote a post on that very point.
MMT Transaction Model - A Variation
(The following is not directed at Dan K.’s articulate comment; it is rather a broad observation.)
I could almost get more value from reading “market monetarist” posts these days than from witnessing the blogosphere train wreck of tortured conceptualizations that has become "MMT”.
MMT has valuable insights into the nature of the monetary system. But it is ineffective as a platform for conceptual exposition of the monetary system. It seems incapable of distinguishing consistently between factual and counterfactual monetary operations. It seems intent on conflating these two parallel modes of analysis, with such stuff as “the government neither has nor doesn’t have money”.
It is not logically possible to present any interpretation of a monetary system that does not reference explicit institutional design assumptions. Monetary systems don’t exist without specific institutional design. In this regard, there are facts of actual prevailing design, and there are counterfactuals, and there are differences between those two things.
It is certainly possible to design an institutional monetary configuration in which “the government neither has nor doesn’t have money”. But the existing system as it is designed does not have this property.
I’d love to see MMT turned upside down in its expositional approach. But the MMT’ers are a small group, with a thoughtful investment in their chosen presentation, so I don’t expect this sort of change to happen. And understandably, like most of us, they probably don’t appreciate criticism at a fundamental level.
"But the existing system as it is designed does not have this property."
It does when you consolidate the balance sheet of the government sector and 'zoom out'.
It very much depends what level of abstraction you are working on at the time.
I do this all the time when designing systems. Sometimes I'm zoomed out ignoring the specifics, and sometimes I'm zoomed in dealing with the nitty gritty - often below the level you talk about (how do transaction records get from A to B in a timely and secure fashion?).
Sticking at one level of abstraction, or constraining yourself by the 'current design' is a huge mistake.
What we can learn from the 'current design' we largely have. Now to explore what the new design should be to deliver the required goals.
Neil,
I'm aware of the abstraction. It's not a question of constraining one's view. It's about being clear on the starting facts.
If you consolidate all of the balance sheets in the world, what you end up with is a single balance sheet. On the left is all of the real assets of the world. On the right is a global net worth valuation of them.
All financial claims net out in such a consolidated view. But the conclusion from that is not that I "neither have nor don't have money".
Such a conclusion would be the height of silliness (unless you assume a design change to barter).
Yet it's the same point.
On the left is all of the real assets of the world. On the right is a global net worth valuation of them.
In what denomination? Martian Credits?
That matters.
Which is why when you consolidate the balance sheet of a sector containing the currency issuer (which is the currency the balance sheet is denominated in) then you don't have any money.
You can't because the accounts cancel out and vanish.
This is why Warren avoids the term 'money'. It's too much of a Humpty Dumpty word.
I wish he'd do the same with 'savings'
"In what denomination?"
Whatever you want, at whatever the exchange rates are. The choice is irrelevant to the point I'm making, which is that the neither/nor mantra is nonsense in the context of the facts that underlie the consolidation.
Far from me to defend his use of particular words, but out of curiosity, what's the issue with 'savings'?
(I agree 'money' is often hopeless due to imprecision.)
Overloaded. net-savings, savings desires, etc all of which refer to different formulaic constructions, even before you get onto the real/nominal split.
'inflation' is similar. Anything where you have to start the debate 'what exactly do you mean by x'.
agreed
> "I could almost get more value from reading “market monetarist” posts these days than from witnessing the blogosphere train wreck of tortured conceptualizations that has become "MMT”."
> "And understandably, like most of us, they probably don’t appreciate criticism at a fundamental level."
Hint: probably easier if you skip the insults.
Hint:
Insults apply to people, not ideas.
Ah, there you go, full steam ahead then with the criticism! Train wreck is waiting for you. Sounds like it's going to work really well.
come on, guys. Can't a fellow take a day off for Thanksgiving without a comments thread degenerating into invective?
JKH: I was wondering if that comment of mine would get you upset. You are, of course, 100% correct and it is *not* accurate to say that the Government doesn't have or not have money, at least under today's operational realities.
In my defense, I've found that the most challenging part of understanding MMT is not that it's complicated--it isn't--it's that you have to unlearn so much other stuff first. I hate the word "paradigm" but Mosler is correct, you are either in paradigm or you are not.
To me, it is easier to switch paradigm via blunt (but inaccurate statements) like "the Government doesn't have or not have money" first and then understand the actual details later, than to try to slide into paradigm by assimilating details of whatever crazy accounts the Treasury uses to buffer tax collection and reserves etc. But, YMMV and my path does involve falsehood--or if you prefer, unforgivable exaggeration.
NEIL: JKH is correct in what he says, and JP's point is correct. We can discuss what would happen if push comes to shove and the Treasury is bare -- will the Fed really bounce a Treasury check? -- but at a formal (I would say pedantic) level, current institutional arrangements do not erase the line between the central bank and Treasury the way they are erased in MMT sectoral balance analysis.
OHM: Government borrowing is changing the term structure of outstanding liabilities. It also drains reserve accounts which is necessary for Fed open market operations (which, before OIR, set the FFR)
WSP,
Thanks for your even keeled response. I should stop this, but my frustration is with what I see as a body of work about a very interesting subject (one that is close to my heart and experience), but that seems to me to be imploding on itself instead of growing as a framework for understanding monetary systems generally. For example, notwithstanding the paradigm’s classification (inaccurate in my view) of currency “issuers” and “users”, there is no reason why the US and the Euro zone can’t be analyzed under the same microscope. It’s really not hard to explain that a central bank has the operational capability to provide deposit balances directly to government, or reserve balances to banks, as an alternative to the government issuing bonds, and that this is an operational option whatever the political linkage between the bank and its non-bank government client(s). And it shouldn’t be difficult to see that this is not the normal modus operandi in any example, but that it can be invoked by those with the political will if necessary. And the difference between the case of the Fed and the ECB is only one of political difficulty. The currency issuer is always the central bank, not some non-bank part of government. And this should be an acknowledged fact that is independent of the degree of political linkage between the bank and the non-bank parts of government.
(BTW, Auerback’s explanations of EZ operations and options tend to be pretty darn good. I'd actually use his interpretation of the ECB as currency issuer as the template for the explanation of what it means to be a currency issuer. He is one of MMT's stronger spokespeople, IMO.)
I need to stop this. Thanks again.
JKH: I think that's an excellent way to put things. The Federal Reserve is the currency issuer, not the Federal Government.
To what degree those two entities are truly distinct is a worthy topic, but fundamentally political, not economic.
"The currency issuer is always the central bank, not some non-bank part of government. "
In the current design. But it doesn't need to be and there is a strong argument that it shouldn't exist at all. The counter-party should just be the Treasury.
"but at a formal (I would say pedantic) level, current institutional arrangements do not erase the line between the central bank and Treasury the way they are erased in MMT sectoral balance analysis."
I disagree. The US institutional arrangements are not the world's arrangements. In the UK, for example, the central bank *is* owned 100% by the Government. Therefore as a matter of standard accounting construction if you consolidate the balance sheet, it disappears.
And again current design, and the political attempt to privatise the currency issuer, should not impede thinking about how it works fundamentally.
Constraining the fundamental model to what can be done practically in the near term is the step that moves the economic ideas into the realm of politics.
Neil,
"In the current design. But it doesn't need to be and there is a strong argument that it shouldn't exist at all. The counter-party should just be the Treasury."
Right - and MMT should state its position just the way you did, Neil. That’s my point.
"Should" is a form of counterfactual, and you’ve implied a very specific institutional design alternative associated with that.
That alternative institution incorporates both treasury and banking characteristics. The treasury characteristic is that it is basically a mismatched balance sheet that “funds” the deficit – the equity position is negative. The banking characteristic is that it is an issuer of currency and bank reserves. And there is no government deposit account on the liability side of this balance sheet. (The previously existing positive central bank capital position also disappears.) Only with such a specific counterfactual institution and such a balance sheet is it correct to say that the government “neither has nor doesn’t have” money. It is clear and unambiguous what is being referred to in this case.
My point is about clarity, consistency, and communication. I think it would help a lot in MMT community discussions, but that’s just me.
I’m also not naive, and I know that MMT head office will never make such a change to the expression and exposition of their paradigm.
WSP,
The market constantly evaluates the political contingencies that can determine institutional configurations and operations. That is a part of risk and therefore a part of the related economics. It’s happening right now in Europe in an obvious way. And it happens implicitly in the case of the US, where the market attaches some sort of longer term probability to the contingency that the Fed would step in if the US had problems with its bond auctions at some stage. MMT itself makes absolutely no sense without such an implicit assumption around standard fiat currency regimes, beyond the ambiguities I’ve described that are associated with it.
One of those contingencies is the near zero probability that the treasury function and the central bank might be combined into a unified institutional currency issuer at some point. That is certainly not remotely on the radar screens of bond investors, although the more conventional alternative of extended bond buying and related reserve expansion certainly is.
It's strange because I would have thought that rolling the Central Bank into the Treasury would have political legs at the moment.
I would have thought it's be a relatively easy sell - bringing the banks under democratic control, money for the benefit of the people not the banks, etc.
An interesting article by Randy seems to go some of the way to making the point.
From my perspective, as a citizen of a country where the central bank is most certainly part of the government in all senses, it seems strange that there is so much reluctance in the US to collapse the institution into the Federal government.
Is it a state thing?
@winterspeak
> "We can discuss what would happen if push comes to shove and the Treasury is bare -- will the Fed really bounce a Treasury check? -- but at a formal (I would say pedantic) level, current institutional arrangements do not erase the line between the central bank and Treasury the way they are erased in MMT sectoral balance analysis."
> "The Federal Reserve is the currency issuer, not the Federal Government."
> "To what degree those two entities are truly distinct is a worthy topic, but fundamentally political, not economic."
Don't really want to come out as the train wreck here, but I like the consolidation abstraction of sovereign currency issuing government. I think it adds value, as it removes a bit of complexity. In fact, as a non-economist, I find the abstraction highly intuitive and revealing.
True, if this removed complexity is crucial for the analysis, then it's a bad idea. Personally, I have not -- yet -- understood why this would be the case though?
JKH remarks: "My point is about clarity, consistency, and communication. I think it would help a lot in MMT community discussions" -- and that I won't disagree with. (And I would like to help improving, but need to learn first. Need to be more clear when we talk "abstraction" and when we talk "specific case", for example?)
But I'm not sure if the consolidation abstraction actually leads to erroneous conclusions? What crucial aspects of the analysis get lost due to the abstraction?
Questions on the case of the U.S (which I don't have much knowledge about):
* Is not the Fed ultimately a "creature of Congress"? (Tried to Google a bit, and found this informative but horrible piece -- http://www.house.gov/jec/fed/fed/fed-impt.htm -- "This paper provides a brief overview of what Members of Congress should know about the Federal Reserve")
* If the Fed bounces Treasury checks, could that not be seen as a spending obstacle that is self-imposed by the government? Would that somehow refute MMT theory? Seems to me MMTers are on an on about these self-imposed constraints?
Neil,
The near zero probability is my personal view. You can make the case for a higher one based on the political heat at the moment. And I've said many times that MMT's general approach is consistent with institutional unification.
Hugo,
I don't think it changes conclusions much, but it would improve the analysis behind those conclusions and the communication of them, in my view.
Hugo:
“Is not the Fed ultimately a "creature of Congress.”
Yes.
So Congress always has the option of changing the nature of the Fed as an institution or the rules under which it operates.
I think it clarifies things to see such potential changes treated as explicit institutional and operational contingencies, separate from current operations, and not assumed away in a soft consolidation interpretation.
Explicit examples of operational contingencies include expansion of understood CB authority to buy government bonds, either in the secondary or primary markets.
Europe is now debating these kinds of contingencies for ECB operations in a very real way.
Loosening up on overdraft restrictions pertaining to the Treasury account at the central bank would be a more radical change in operations.
Explicit example of institutional contingencies would be a formal fusion of Treasury and the CB operations and balance sheets into a single institution.
Hugo,
“Seems to me MMTers are on and on about these self-imposed constraints?”
My impression is that the MMT model derives a “base case” by assuming (implicitly or explicitly) the elimination of such self-imposed constraints from the actual monetary system we now have. The consolidation meme tends to describe things implicitly as a reflection of that base case. On the other hand, Scott Fullwiler speaks explicitly of that sort of base case for monetary design.
I find the specification of a “base case” which is itself the result of a backward structural derivation to be awkward logically. I’d prefer a model that starts with current operations, and treats adjustments to current constraints, operations, and institutions exactly for what they are – hypothesized changes to the current system. One can then deal with the economic consequences of the system we have in terms of probabilities for the feasibility of such hypothesized changes should they become desirable or necessary.
Example: is it reasonable to think that IF the US somehow reached a Debt/GDP ratio of 120 per cent, and IF the bond market vigilantes returned in full force, WOULD the Fed be lending a big hand by buying treasuries in size? I think it’s reasonable to assume so. And that probability may be different than what it now is for the case of the Euro zone and the ECB.
JKH,
I agree that it would be more helpful to walk backwards.
"I don't think it changes conclusions much, but it would improve the analysis behind those conclusions and the communication of them, in my view."
Would you agree that it doesn't really change the conclusions as long as we make the assumption that the market perceives US debt as nearly risk-free (Congress does not threaten to not make payments)? This is Fullwiler's 'semi-strong form' deficit, wherein sufficient investors and financing with an eye towards arbitrage against current/expected FFR ensures tsy debt auctions succeed (and the interest rate on tsy debt mostly reflects FFR expectations).
That being said, though the end result of 'strong form' and 'semi-strong form' are the same, the process and requirements (sufficient investors, financing, sophisticated capital markets) are different. Even Fullwiler himself says 'financing' is actually required to make a tsy debt auction succeed! (Might I add, in the discussion of the MMT head office, Fullwiler is pretty explicit about how things work in gory detail). Like you, I wonder why Mosler, an arbitrageur himself, is so comfortable lecturing as if the Treasury simply just spends, and is so adamant that the funds the Treasury spends are what buys the tsys.
" is it reasonable to think that IF the US somehow reached a Debt/GDP ratio of 120 per cent, and IF the bond market vigilantes returned in full force, WOULD the Fed be lending a big hand by buying treasuries in size?"
And so I just want to add, 'bond market vigilantes' would only return if 1) markets were either irrational and/or 2) there was political uncertainty regarding Congress's willingness to make payments (ie, not raise the debt ceiling). JKH, do you agree?
"I disagree. The US institutional arrangements are not the world's arrangements. In the UK, for example, the central bank *is* owned 100% by the Government."
Even with the BoE, the old way by which it could lend directly to government - via its "ways and means" advance/overdraft facility - has been effectively neutered. That facility was frozen in 1997 and has since been almost completely repaid. The upshot is that MMT can't look to the BoE as an example of its idealized "consolidation", and it can't give policy recommendations as if these institutional rigidities didn't exist.
I think such obsessive focus on the 'institutional arrangements' is misguided. MMT doesn't rely on a consolidated view, even though that's the view many of its proponents like to take. I think this is partly Fullwiler's message- ****It's the interest rate, stupid.****
Economically, it just doesn't matter if there are overdrafts or not.
As long as markets don't seriously believe Congress will prevent payment (even this summer interest rates were not phased!) and are rational, the tsy auctions will succeed and the govt will spend, as if there were no funding constraints and as if there were overdrafts. There's an arbitrage profit to be made if the interest rate on govt debt does not reflect the FFR.
End of story.
Anyone disagree?
Problem is often semantic. For example, the cb is the currency issuer in the sense of issuing reserves (banks obtain cb notes in exchange for reserves). Reserves are never actually spent in the economy, since they are used only for settlement and reside on the cb spreadsheet. But most people think of currency in terms of something they can spend. Reserves obviously don't qualify in such terms. In this sense, bank reserves on the FRS spreadsheet are not what most Americans consider "dollars," for example.
Of course, most people have no idea that all of this "money" is created "out of thin air" by bank lending or government expenditure. The president is fond of saying, "Money doesn't grow on trees." No, it just comes out of thin air, but only if one has access to the magic wand.
What I see as a major obstacle in the promulgation of MMT is the semantic difficulty of bridging the gap between the technical meaning of terms and the ordinary language meaning that most people are going to be thinking in terms of.
I think that the MMT economists have attempted to be technically correct in communicating with their peers, but maintaining this degree of precision with others makes the message impenetrable. As soon as one attempts to state the matter precisely in imprecise, understandable terms, some of the precision is lost. But if one tries to use profession jargon precisely in communicating with the public, then the communication gap becomes insurmountable for many if not most.
Of course, this semantic difficulty isn't something that affects only MMT. It's pretty common in other field. Student may not be able to get to the nitty gritty of technical precision until they reach the PhD level in some fields. I'm not sure there is a solution to this in any field in which complex concepts of a technical nature are involved. Certainly the problem is exacerbated in economics and finance, where many terms are ambiguous, understood in one way by professionals and in a very different way by others.
Maybe instead of complaining we should all be cooperating on getting this message boiled down to something that non-technical people can grasp before the world implodes.
wh10:
“And so I just want to add, 'bond market vigilantes' would only return if 1) markets were either irrational and/or 2) there was political uncertainty regarding Congress's willingness to make payments (ie, not raise the debt ceiling). JKH, do you agree?”
No.
The fact is that the term “bond market vigilantes” arose from observation of actual behaviour of bond markets in the past. That behaviour was associated primarily with perceptions of inflation risk and consequent market pricing of that risk into bond yields. And that was in turn supported by market expectations of a requirement for the Fed to raise interest rates in response to that inflation risk. The market was pricing in an arbitrage relationship that it expected would be the rational consequence of required Fed policy. So the “vigilante” term is strongly associated with the bond market “front running” the Fed in the sense of attempting to force onto the Fed its own requirements for Fed policy.
There is nothing in the above that has to do directly with solvency.
What I hypothesized was similar behavior in the context of a much higher debt/GDP ratio than we’ve seen since the term “bond market vigilante” was invented.
So neither of your qualifications is necessary for the scenario I described.
But the combination of such observed vigilante behaviour and a very high debt/GDP ratio might lead to some interesting dynamics that we haven’t seen before, in terms of institutional adjustment as I suggested.
And my point is not about the probability of the occurrence of the sort of scenario I described. It's about explaining what it would mean if it happened.
Tom,
“We should all be cooperating on getting this message boiled down to something that non-technical people can grasp before the world implodes”
I would substitute “most” for “all”. In crude terms, MMT consists of a technical core understanding, plus the marketing or communication of that understanding (and the associated policy options) to various audiences, including those who are looking for the distilled version.
“All” in your prescription above implies that this technical core understanding is perfectly formulated, and incapable of further improvement. I wouldn’t share that view, but that’s just me.
I think it’s probable that your prescription is true for the majority of those who have a serious interest in MMT. And more power to their communication efforts.
On the other piece, “core understanding” is a self-referencing idea, in that the MMT principle founders/leaders are the ultimate arbiters of the thought construct. So I don’t expect my occasional wheel spinning on this issue to have much influence. But I don’t see that it should hurt.
@JKH,
Your comments are very helpful. (I refer mostly now to the two marked "7.24AM" and "7.46AM").
A couple of excerpts (merged):
> So Congress always has the option of changing the nature of the Fed as an institution or the rules under which it operates.
> I think it clarifies things to see such potential changes treated as explicit institutional and operational contingencies, separate from current operations, and not assumed away in a soft consolidation interpretation.
> My impression is that the MMT model derives a “base case” by assuming (implicitly or explicitly) the elimination of such self-imposed constraints from the actual monetary system we now have. The consolidation meme tends to describe things implicitly as a reflection of that base case.
I think I understand your points. I hadn't thought of it, but yes. It seems the "base case" (or "general case") is a case where self-imposed constraints are assumed away. (Or at least assumed to be "reasonably easily modifiable" or something).
If one would have accepted that assumption (something you are reluctant to do, right?) the consolidation makes more sense, right? (Just to check if I understand your message).
Personally, I'll try to be more explicit about this.
However, I think I am actually still in favor of using this consolidated-government-abstraction. My defense would be:
* This "base case" is about one order of magnitude less complex to describe, in my opinion, which makes it a suitable starting point for pedagogical reasons if nothing else.
* It seems consistent with a theoretical Chartalist view of fiat money (as an IOU-token issued by some sovereign body of authority, and in modern instances driven by taxing). This view has sound academic basis, as far as I can tell.
* This view also explicitly juxtaposes "self-imposed" constraints from the actual "operational" constraints (lacking a better word) that a sovereign currency issuing government actually has to abide by. This can be a fruitful distinction to make. To clearly understand and spell out the options actually available can be empowering.
* The "self-imposed" constraints should be acknowledged as well of course, and can be understood as "layered" on top of the "base case". (Personally, again, I should try to convey this better in the future.)
* The "self-imposed" constraints should perhaps be thought of as belonging to the realm of political sciences(?), rather than economics? (Not sure it is possible to make such distinctions at all times, but I have a vague feeling it could be useful in this case.)
Also, I thank you for being so patient with us, in spite of any frustration you may feel on these issues.
"Even with the BoE, the old way by which it could lend directly to government - via its "ways and means" advance/overdraft facility - has been effectively neutered."
Also, Bank of England lending to the UK Treasury not allowed by the Maastricht Treaty ;-)
Hugo,
I think I agree roughly with your last comment.
Further elaboration on my take in the next comment.
Hugo,
The most relevant “self-imposed constraints” (SIC) in my view are:
a) CB purchases of government bonds in the secondary market
b) CB purchases of government bonds in the primary market
c) Direct credits/overdrafts to the Treasury (TR) deposit account at the CB
d) Treasury Bond issuance
The first SIC relates to the freedom of the CB to purchase bonds in what is effectively a QE operation. This is permissible under prevailing rules, but an ethos of policy restraint exists around it. The next two relate to outright prohibition of CB direct credit support to TR, both asset and liability sides – no credit by bond purchase, no deposit/overdraft credit.
Any of these could be relaxed in the context of the current institutional separation of TR and CB functions. It requires operational adjustment, but not fundamental institutional realignment.
The final SIC is bond issuance. This constraint could be lifted, but is an institutional game changer. With it, the first three constraints become moot and eliminated de facto (see below). I would define this as the “base case”.
The “base case” is only consistent with institutional reformation, as follows:
Elimination of bonds means that the central banking function provides for the continuous absorption of government payment and receipt activity into the stock of bank reserves. The combination of bank reserves and currency grows in conjunction with net deficit spending.
Because there is no longer any bond issuance, there is no longer any operational requirement for a Treasury deposit account at the Federal Reserve. The flow of Treasury payments and receipts is simply absorbed into the central bank reserve position.
The Treasury deposit account at the Fed reflects the institutional separation of Treasury and Fed balance sheets. Because there is no longer a need for that account, there is no longer any need for such separation.
So the two are fused and replaced by a single balance sheet corresponding to a single institution.
Working from this balance sheet, one can more accurately say that the government “neither has nor doesn’t have money”. The reason for that is that Treasury has operationally taken over the role of currency issuer. Treasury no longer is constrained to currency user status in the sense of needing a deposit account at a separate Fed. Treasury incorporates the banking function as principal, and issues and retires reserves directly as a marginal reflection of its own cash flow profile, which now includes both budgetary flows and any other central banking activity. Budgetary payments and receipts expand and contract the balance sheet like normal CB OMO.
Logic:
The existing monetary system with its constraints is the factual one. The intermediate and base cases are counterfactuals. There are two different modes for converting counterfactuals to factual systems:
a) Permanent
b) Temporary
Given the comprehensive nature of the base case, any change to it would likely be intended as permanent. Some MMT’ers favour this.
A different type of change would be temporary, as in an emergency. This would occur more likely in the mode of the first three types of operational change, without fusing CB with TR.
I think the key to the interpretation of US fiat solvency is the market’s implicit understanding that the US government can authorize permanent or temporary change to US monetary architecture. This becomes reflexively reassuring - actual change won’t be necessary, given the market’s confidence in this capacity for change.
This is not the case in the Euro zone, where political difficulty in securing such changes is treated as higher risk than likely would be the case in the US. But the menu for actual technocratic change is essentially the same, as it reflects basic operating arrangements between the central bank and its treasury function(s).
JKH,
“I think the key to the interpretation of US fiat solvency is the market’s implicit understanding that the US government can authorize permanent or temporary change to US monetary architecture. This becomes reflexively reassuring - actual change won’t be necessary, given the market’s confidence in this capacity for change.”
I am not sure I agree 100%, but I might and would like to hear your thoughts. This is what I was trying to get at with my comments above, but they got derailed since I misinterpreted what you were addressing at the time.
For the purposes of fleshing this out, let’s assume away the possibility that Congress will block debt issuance, and more importantly, that the market doesn’t fear this.
In this world, the arbitrage relationship between govt bond yields and current/expected FFRs will ensure tsy debt auctions will always succeed and that the US stays solvent. You need sufficient investors and financing for this to occur, but I see no *realistic* reason why these wouldn’t exist outside of the country collapsing (I am willing to be corrected here). The ‘reflexive reassurance’ may be a real force in this world, but if it were, I would argue it’s reflective of an irrational market, in the sense that they don’t understand monetary operations and US govt debt markets. I am trying to describe, more or less, Fullwiler’s ‘semi-strong form’ of deficit spending.
If, instead, we allow for serious political uncertainty regarding debt issuance, then I think the ‘reflexive reassurance’ becomes a more ‘rational’ factor.
However, in today’s world, I wonder how important the ‘reflexive reassurance’ factor really is. This summer, US govt bond yields did not seemed phase much by the debt ceiling debacle, and the market certainly wasn’t pricing in, at least in the short run, the possibility that Congress changes the monetary architecture (that the US was a safe haven during this time period may be a confounding factor). So I would argue, today, and likely in most realistic scenarios, the ‘semi-strong form’ holds, and it is the Fed’s ability to set the FFR and the arbitrage opportunity it presents that is the key to understanding US solvency. The possibility of changing the monetary structure is not necessary.
I discussed this recently with Fullwiler here. Hopefully I didn’t misinterpret his message, and as I said, I would be curious to hear your thoughts. http://heteconomist.com/?p=2835&cpage=2#comments
Won't any government managing a fiat currency need to issue bonds simply to preserve the incentive of currency users to hold and trade in that currency? Suppose we think of bonds in MMT fashion as the equivalent of savings accounts, and bond purchases and redemptions as the equivalent of moving money from checking deposits to savings deposits and back again. It seems to me that a potential currency user will be reluctant to hold a currency that doesn't offer an interest bearing savings vehicle. They will seek out a currency that does offer such a vehicle, and convert their savings into and out of that currency as needed. If the transaction costs are low, this might not be that inconvenient to the currency user. But the government would lose control of monetary policy.
I was also wondering about the possibility of an intermediate policy tool for managing price stability. MMT encourages us to see the fundamental role of taxation as a tool for price stability. Taxation removes net financial assets from the private sector while spending adds net financial assets to the private sector. Borrowing also removes net financial assets from the private sector, but it restores those assets later, in full and with interest, and that tool differs from taxation in that it is both voluntary and temporary.
What I'm wondering is why there isn't a third kind of tool which is temporary like bond issuance, but which is mandatory and pays negative interest. For example, one might receive a tax bill for $10,000, but upon the the payment of the tax the taxpayer might receive a coupon that can be redeemed after five years, for example, for some pre-determined amount - say $2000. Yields and durations of these partially rebated taxes could be adjusted from period to period as a part of monetary policy.
Govt borrowing doesn't remove NFA. It replaces one with another.
wh10,
I regard the debt ceiling episode as well as the S&P downgrade as mild occurrences in the context of potentially more severe scenarios such as the specific one I alluded to earlier. Also, the construction of the latter was quite consistent with the idea of arbitrage between expected Fed policy and the yield curve. In any event, expectations regarding potential operational or institutional adjustments are not immediately observable. And all that said, various pundits and Geithner himself were quite vocal about the possibility of delayed payments, which in some scenarios might be a prelude to other operational changes.
I found myself agreeing with much of Lavoie's paper when I first read it. At the same time, I'd be interested to read Fullwiler's concerns about it. I hadn't noticed the extended discussion you linked to, so I'll have a look.
DAN: No, a Govt just needs to be able to tax in a currency for there to be demand.
Govt "borrowing" does not remove NFA from the private sector. It just changes their term structure.
JKH,
Looking forward to it. Fullwiler's thoughts are there. To me, he is saying the architecture debate misses the forest for the trees. It is all about the Ffr and arbitrage. Like you said, your bond vigilante example has nothing to do with solvency, but inflationary expectations; and so any changes to the architecture in that situation still wouldn't have to do with solvency. In any case, with the yield curve on the Fed's leash, the Fed could deal with those vigilantes without resorting to architectural changes.
But if we are talking about solvency, there seems to be a catch 22 regarding the potential for changing the architecture. Again, in the US, assuming sufficient investors and capital markets, there should only be solvency concerns if there is political uncertainty regarding Congress allowing the US govt to issue debt. But if that uncertainty exists, then why would an investor believe Congress would be willing vote through a new monetary architecture that allows the Fed/Treasury to spend directly? From this perspective, changing the monetary architecture seems even less relevant.
So, as I am arguing, the key to US solvency seems to be more the ability to set the FFR and arbitrage, not the possible existence of a monetary architecture with less self-imposed constraints.
wh10,
Just reading your comment again quickly, I think we're in rough agreement. What I called "reflexive reassurance" wouldn't be much of a critical factor other than occasions of perceived extreme risk, such as the scenario I painted earlier, and which was a scenario primarily of perceived interest rate risk in Fed policy. We're a long way away from that. The debt ceiling impasse was perceived as low risk in the sense that the market sensibly anticipated that something would be done to fix the operational glitch of a redundant debt limit, one way or another. The market more sensibly interpreted the whole thing as a Congress that was generally dysfunctional and specifically dysfunctional in the case of coming to grips with the correct fiscal response to deflationary pressures - so bonds continued their journey in the direction of zero rates.
wh10,
I agree with Fullwiler that the primary issue is interest rates. The contingency of changing operations and/or institutional arrangements would be perceived by investors as the capacity of the Fed to abandon the market’s pricing of term debt and set the interest rate directly on reserves instead – short term or whatever term the Fed desires. That’s an interest rate choice, not a solvency choice. I don’t think I’ve put forward a rationale about solvency per se, at least as it pertained to the example I was using for illustration purposes. The debt ceiling thing is a little different in that it’s just a dumb ass “risk limit” that could potentially get in the way of continuous government transaction activity for a few days.
Before I go back to the Lavoie paper and that link, I’d just say that my recollection of it as well as what I’ve written here is that the issue being addressed has to do with the presentation of the framework for the analysis of monetary operations. It’s primarily a style issue.
JKH,
Please see my above post and the following. Can you clarify what your imagined scenario would entail? What would be the perceived extreme risk exactly? Are you talking about a scenario where we are approaching hyperinflation, so investors do not believe taking advantage of the arbitrage relationship is worth it since the currency is viewed to be worthless and thus auctions fail? Outside of that, I am curious to understand what other risks could be rationally perceived. Perhaps I am too naive or wrong in believing in the power of bank lending (see step 1 in Lavoie's post-chartist table 3 accounting) or in investors to arbitrage away profits. And if you are talking about being another debt ceiling debacle, then I see the catch 22 issue mentioned above.
This comment has been removed by the author.
Ah, sorry, posted before seeing your last comment.
I think I agree with the first half of your first paragraph, but do we not also agree that the Fed currently has all the capacity it needs to set the price of term debt anywhere along the term structure?
I agree it is a style issue, but your comments in the blogosphere have earned much respect from me, and so I am uncomfortable regarding your statement above that the key to understanding US solvency is the possibility of changing the architecture. I don't think this is right, and if this was your position, then it goes beyond a style issue. But it seems we now (roughly) agree.
As someone who's been an enthusiastic MMT explainer since early 2007, I find myself agreeing with objections several people point out:
1) "Govt neither has nor doesn't have money." This simplification creates more confusion then it clears up. The use of the term money is problematic, the use of the word govt is problematic (are we talking fed+tsy or not?) and it's not even true.
2) Use of the term "printing money." Even I violate this one sometimes. Better to just be specific about whatever action you're talking about, usually the fed purchasing tsy secs.
3) "Govt spends by crediting accounts/keystrokes" It leaves people feeling very unsettled because they know bond sales must match deficit spending, since you're teaching MMT you should include the self imposed constraints the country you are describing has.
4) "Government is the monopoly issuer of the currency." followed by "private bank loans create deposits." Lucy, you have some splaing to do.
5) MMT's use of the word leverage. It's correct in the way accountants use it, but unfortunately the rest of the economics profession interprets it like the money multiplier.
5a) For me, and I'm sure a lot of people are in the same boat, they find themselves hitting a language barrier when communicating with other economists. It's good to use words as they are defined and if you want to be successful at communicating your ideas add "in other words" following words and concepts that are prone to being misunderstood.
MMT has yet to "win" any major economist (that I'm aware of). In the meantime, why not invite Keen, Quiggen and other intellectually honest economist to a weekend of trying to reach common understanding?
I love MMT, I have nothing but the highest respect for the MMT academics, I'm writing this in hopes it can make them more effective.
Tschaff, I like Fullwiler's brand of MMT the most. There's no verbiage there. It's just straight up accounting and financial markets acumen. Stylistically, Fullwiler's 'strong' to 'weak' framework is the best I have seen; yet I didn't realize that until a week ago (after many many months) since I was lost in and taking for granted every other MMTers' conceptualizations as if we truly do operate in the 'strong' form. It really doesn't matter, save for extreme political circumstances, but unless you understand why it doesn't matter, then you don't really get it, IMO. That self-imposed constraints are like tying your shoe-lace doesn't cut it for me as far as understanding why, but I can understand why that approach is taken for those not interested in spending the time or already having the background to understand why.
To me, this can only get so "simple.". It truly is complex stuff if you want to explain it properly, and until you do it properly, you're going to have valid complaints that MMT doesnt entirely make sense.
And MMT has won Galbraith, whom I would consider major. He invokes MMT where it matters- papers and Congressional testimony. He is less aggressive about it in the media sometimes, but he's effectively an MMTer IMO. Even Krugman associated him with MMT.
wh10
I wrote:
"I think the key to the interpretation of US fiat solvency is the market’s implicit understanding that the US government can authorize permanent or temporary change to US monetary architecture."
It's getting late and maybe I'm not being perfectly consistent. First, by "architecture", I mean the broad swath of operational adjustments and institutional arrangements. Second, I did refer to solvency directly here rather than interest rates. Although, it is possible for bond auctions to fail, as Fullwiler has pointed out. So technically, the ability to come up with the desired cash (and at non-chaotic rates) could conceivably require some adjustments in extreme stress scenarios. I also noticed again that Scott often refers to the example of issuing short date bills rather than bonds, which is another type of adjustment and less radical than reserve issuance through overdraft or whatever.
I skimmed over Scott's comments at the other place and that all looks pretty good to me.
Again, we're talking mostly about points of style in presentation. I once said/wrote to Scott that I would present the monetary system framework in a different way, and maybe I've started to do that a bit here.
And I just noticed that things got pretty rambunctious over at the Mike Norman thread on the same topic. What we're discussing here is mostly about style - not about deficit policy, etc. There shouldn't be any threat to the policy part of MMT from a discussion about stylistic presentation on the monetary system side of things. And I think some people may be underestimating Lavoie's understanding of MMT's approach to monetary analysis, mistaking some of his style criticism for substance criticism.
If Galbraith were a Venn diagram, MMT would be a reasonable size circle within a larger one.
Agree on Galbraith.
I guess I would like to better understand what exactly the extreme stress scenarios would be, outside of Congress threatening to block debt issuance, which to me doesnt count because like you said it is 'dumb ass' but more importantly because the mkt wld never price in the possibility of an architecture change in that scenario. Then I'll better understand why the monetary architecture might matter as it pertains to solvency.
Obviously it counts in that a change would have avoided it, but doesn't count if we are talking about the market expecting a change as a factor in ensuring US solvency. And obviously, monetary architecture definitely matters if you are not at least as close as the US or Japan to 'no constraints' vs say the Euro or gold std or fixed fx.
Just assume as a scenario US financials that resemble Greece - 120 per cent debt/GDP ratio or whatever.
(In my earlier scenario, I assumed high inflation risk premiums in interest rates, but I'm not sure that's even necessary for the example.)
Now assume that the market expects that Congress/ Treasury/ Fed will never make any operational change.
That means the market interprets Treasury to be in the same position with respect to the Fed that Greece is in with respect to the ECB.
And because I've assumed comparable financials, there's no reason to assume different interest rate or bond market behavior. And everybody believes Greece is insolvent.
But all of that is a contradiction with respect to how we believe the market will behave, because we believe solvency is not an issue with the US.
Because its a contradiction, the market must believe that the Fed will be allowed to come to the rescue with required operational change.
And because the market believes that, it won't be necessary for the Fed to actually do that in most cases.
(Proof by contradiction)
- Must go now -
Tschaff,
I too have been a bit skeptical of the usage of the phrase leverage by MMTers.
"MMT's use of the word leverage. It's correct in the way accountants use it, but unfortunately the rest of the economics profession interprets it like the money multiplier. "
I saw Fullwiler commenting on it at Heteconomist as well. Here from Heteconomist
"We are using “leverage” the exact same way that it is used in accounting and financial management—indeed, it’s the same way that all the corporate finance textbooks I use teach it. That is, leverage to NC refers to a leveraging of the balance sheet, as in assets divided by equity—the “equity multiplier” or “leverage” (my students have to learn that this measure is called “leverage” since that’s what the business simulation we use calls it, for instance). In other words, “leverage” here has nothing at all to do with leveraging reserves as in the money multiplier, and it has nothing at all to do with suggesting that reserves or anything else are a priori necessary for leverage to occur."
I think he is defending a wrong thing. There are dozens of places where MMTers say "leveraging of HPM", as opposed to the phrase used in finance and economics.
wh10,
Sorry what is the arbitrage?
Buying a long term bond is not arbitrage even if the yield is high. Because the yield can go even higher (and hence the price lower).
"Even with the BoE, the old way by which it could lend directly to government - via its "ways and means" advance/overdraft facility - has been effectively neutered."
Nope. The ways and means account was used fully only a couple of years ago during the crisis.
"Also, Bank of England lending to the UK Treasury not allowed by the Maastricht Treaty ;-)"
I have a useful note from one of the investment banks that has researched the legislation. Their conclusion is that there is no restriction on the UK government instructing the Bank of England.
Their conclusions on Maastricht are:
"The treaty functions as a guideline broadly defining good behaviour. It is not enforceable"
"So the EU rules prevent nothing"
Once again Ramanan the evidence is against you.
Neil Wilson,
Well that's just your opinion, man.
An advice: If you want to debate with me, do not write lines like the following.
"Once again Ramanan the evidence is against you."
It looks like an act of desperation.
What evidence? What you quoted is just an opinion. Also, the UK has signed the Maastricht Treaty but retaining the option of joining the Eurozone. Hence the ESCB (which includes the Bank of England) is bound by it.
From the Bank of England:
"There has been a change in the composition of Issue Department assets. The Ways and Means advance to HMTreasury was £7.4 billion at 29 February 2008 (2006/07: £13.4 billion). The Ways and Means facility is the central government’s overdraft facility at the Bank of
England. Until 2000, when the Debt Management Office took responsibility for the government’s cash management function from the Bank of England, the facility was used as
the means of balancing the government’s day-to-day cash needs. Since 2000 its size had effectively been unchanged. It has always been the government’s intention to reduce the advance over time. There were two such repayments in 2007/08, of £4 billion on 24 January 2008 and of £2 billion on 31 January 2008. The timing of this action was in order to provide the Bank with additional flexibility in managing its balance sheet. There was a further part repayment of the Ways and Means facility of £7 billion after year end, on 17 April 2008."
So if anything, the change in the ways and means advance (in the balance sheet) was to reduce the advances rather than increase contrary to what you are saying!!!
More correctly, the rule is that Ways and Means advance is limited which is equivalent to a no-overdraft law in most senses.
Please check Table 6 here http://www.bis.org/publ/bppdf/bispap20d.pdf
for the precise statement. It says prohibited by Maastricht Treaty.
It is a different matter whether I am recommending it stay as such and discussions such as that. But if you make an attempt to mislead, you will always be stopped at this point and will never make a progress.
It is true that if Britain runs into problems, they can give violate the Treaty unilaterally or tell the EU, that it is withdrawing, or use such means.
But ... do not mislead!
"What evidence?"
That it has never been enforced.
" It says prohibited by Maastricht Treaty. "
And who's tanks are going to enforce it?
"More correctly, the rule is that Ways and Means advance is limited which is equivalent to a no-overdraft law in most senses. "
Nope. The Treasury has reserve powers given to it in the Bank of England acts which allows it to direct the Bank of England in any way it chooses "as required in the public interest".
The legislation exists in the UK to allow the Treasury to control the Bank of England completely without any recourse to parliament passing an act to allow it to do so.
So in the UK we have complete government control of the central bank as a matter of legal fact.
"So if anything, the change in the ways and means advance (in the balance sheet) was to reduce the advances rather than increase contrary to what you are saying!!!"
Have a look at this picture
Is there an increase in the ways and means account around January 2009?
So I'm afraid you are mistaken. There is no restriction on the UK government borrowing from the central bank if it feels the need to do so.
The UK could go MMT tomorrow without requiring a single act of parliament.
Neil Wilson,
Get your facts right. Overdrafts limited as per the BIS article.
"And who's tanks are going to enforce it?"
Change of topic.
Ramanan,
The arbitrage is that of the expected policy rate path versus the long rate.
This is unobservable, but the logic is reasonable.
However, the critical point is that the expected policy rate path can and does change on a dime. It is reasonable to interpret high volatility in the expected policy rate path, and therefore high volatility in bond rates, particularly those of longer term. I think that goes to your point.
An unusual commitment to a given policy rate path can reduce that volatility, but only on bonds with a term less than or equal to the commitment period.
"The UK could go MMT tomorrow without requiring a single act of parliament."
So it doesn't run according to "MMT" guidelines?
JKH,
Yes I agree.
My only point was about the phrase "arbitrage"
But my point was that it tends to get overused - so for example, there are some articles about cash management of the US Treasury where worry is about reducing interest paid on these bills.
The comments by wh10 kind of gives the impression (when applied to CMBs) that the cash management bill will always be absorbed at the announced rate.
Ramanan,
Perhaps I am using it too loosely, but what I meant is along the lines of what JKH said - the current/expected path of FFRs. Fullwiler could explain in more detail the various kinds of arbitrage trades (e.g., futures, swaps) that ensure that the LT rate won't vary much from future expected ST rates. I don't know what you mean by your cash mgmt example.
JKH,
I am glad you went there, as I wanted to ask how my logic (which excludes market expectations of a changed architecture) would apply to say US states or Greece.
I think Lavoie made an attempt at explaining why in his critique, though I need to finish reading it. What I think he said was that the ECB is not supposed to buy govt bonds in the secondary markets, which places a systemic tension in the European banking system. That tension doesn't exist for the US banking system as it pertains to federal govt debt, since the Fed does buy in secondary markets and can act as a lender of last resort. But that tension does exist for US state debt, which may explain why US states are like Greece.
If this an adequate explanation, then you still don't need market expectations of architectural change to explain why markets won't behave as they are in Greece with respect to the US. The US architecture is fine as it is. The US should never become Greece, even in your hypothetical situation with similar financials, because the Fed is a buyer in secondary markets and is allowed to be a lender of last resort.
wh10,
When financial analysts and traders talk of "arbitrage" they are essentially talking of trades which tries to exploit price differences between bonds and various fixed income securities and contracts.
One example is exploiting the difference in the the interest rate swap rate and government bond yields.
These trades are far from being arbitrage trades and there are zillions of risks around them.
For example the institution doing this will profit if there is no default in the end (in the swap) and during the lifetime of the trade, the swap spread doesn't widen. i.e., it - the institution has the capacity to hold the trade till maturity.
Also, more importantly these trades are about different rates on the same point in the maturity space. So you could try to make money in Treasury Futures if the price of the Treasury Futures and underlying bonds is out of whack.
There are other trades which try to take a view on the shape of the yield curve - such as steepeners and flatteners. The idea behind them is that the yield difference between two points in the yield curve is either large or small and should change. These are nowhere close to being arbitrage trades and are simply directional bets.
Hence I cribbed about the usage of the word "arbitrage".
The sense you started off it something different.
But its good to not mix the various situations and contexts in which the word is used.
wh10,
See my previous comments. I took pains to qualify my use of the term "architecture" to include operations that deliberately depart from the norm, without necessarily changing the institutional configuration, but including that possibility as a final option. I've been consistent there.
(Granted "architecture" may not have been the ideal word to capture this, but I qualified it.)
QE is such an operation. That's why its a big deal since the crisis.
See my various earlier comments. QE is the first such operation on the list. The actual, comprehensive institutional change that I defined as the "base case" is the last. It's a continuum of potential change, starting with QE and ending with actual TR/CB institutional combination.
And the interpretation of QE as an abnormal operation in the context of the Fed is the same as it is with the ECB. It's a subject of debate on both sides of the Atlantic. The Fed has made its operational change, which everybody acknowledges is an unconventional policy, but the Fed still exercises a degree of "self-imposed constraint" by iterively limiting the size of its program. The ECB can become more like the Fed simply by lifting what amounts to a more severe policy restriction on using this unconventional policy. And interestingly that's something that the MMT'ers have recommended it should do as a minimal adjustment.
P.S.
I meant it’s the same general interpretation regarding the interest rate effect.
The Fed is swapping reserves for bonds to set interest rate sensitivity shorter, and interest rate cost cheaper.
The ECB is at a different stage, which includes fiscal crisis. But it could do the same things as the Fed, in order to set interest rates shorter and cheaper, thereby alleviate debt servicing cost pressures. That would provide “reflexive reassurance” to EZ markets.
JKH
I agree, and you are right- you did qualify this. Do you agree, though, that it is not just the need for QE during abnormal times, but also the need for the CB to swap reserves for sovereign govt bonds on a day to day basis, that will prevent escalation to crises such as these? Again, I am trying to validate Lavoie's point.
Ramanan,
I am aware of what you said but do need to learn more of the technical details. It would be helpful if you could point out where my use of it was quesstionable. I am trying to use it as Fullwiler does. Does he use it improperly?
JKH, if I interpet you correctly, it seems you are essentially saying this. But I guess I am talking more about what tools the CB has to ensure payments settle, and crucial to this is managing reserves adequately, which includes OMO. The intent of QE in the US currently seems different, but it is effectively what is done day to day to manage reserves, make sure payments settle, and defend the FFR.
"Nope. The ways and means account was used fully only a couple of years ago during the crisis."
It was a relatively small operation in the context of the historical scope of ways & means operations, and lasted for only a few weeks. And it doesn't change the fact that the ways & means facility has been effectively neutered relative to what it once was, and that the Treasury has been signaling its intention to reduce its dependence on such advances for over a decade.
Did you know that ways and means advances once accounted for 20-30% of BoE assets? The current balance is some 300mm pounds, a miniscule 0.001% of assets. So the modern day UK Treasury is surely not spending via its ways & means account.
There are no definite poles here, just shades of grey. Pure black is a banking structure that allows for the fully consolidated MMT way of thinking, pure white is a completely independent central bank. In general, most central banks lie nearer to the white pole, and they're getting whiter by the decade, as is certainly the case with the BoE.
"The UK could go MMT tomorrow without requiring a single act of parliament."
Sure, the BoE could start to move back to the black pole. But that's not happening. The BoE is already a medium grey and is only getting whiter.
Anyways, no more comments on ways & means from me.
wh10:
I don't think we're far apart.
wh10,
"This is Fullwiler's 'semi-strong form' deficit, wherein sufficient investors and financing with an eye towards arbitrage against current/expected FFR ensures tsy debt auctions succeed (and the interest rate on tsy debt mostly reflects FFR expectations)."
This one.
In the purest form, an arbitrage argument starts from a portfolio of value 0, when funds can be borrowed to make a trade and the trade pays off in some time interval 0 to T.
Or else, one could frame it differently - say that I have funds of $x and I find I can find a risk free security or a combination which pays me a higher interest in a time period 0 to T greater than another risk-free rate with certainty. Such a trade will be arbitraged away etc.
An institution participating in a government auction has none of these luxuries. For if it is buying a 30-year bond, there is no arbitrage really, strictly. It risks a rise in yields of bonds in the future.
For example, expectations of markets on prices or yields can be self-fulfilling, leading to rise in yields.
A buyer of a 30 year bond at the auctions risks that expectations of the market can go wrong.
Of course, Neoclassical Economics has nothing much to say - it has exogenous money and higher supply of government bonds leads to rising yields in the scenario.
The mistake is of course that it ignores that government deficits increases private sector wealth.
Back to real world... the Treasury and the Central Bank are constantly involved in expectations management. If some segment of the market has excess supply, the Treasury can move its supply to short end of the curve and this has the effect on prices of bonds and indirectly on expectations because as prices move expectations change.
The central bank through its communication is also affecting expectations.
It can also enter secondary markets to change expectations from running away.
(The ECB is doing that, preventing expectations from heading into disaster land, at least attempting to).
So back to the 30y bond buyer, he is not making any arbitrage. He is just making more returns on an average as compared to always investing short term since the yield curve is usually positively sloped. It is misleading to call this arbitrage. For example in 10 years, yields may be very high and he won't be able to sell it with the return originally thought.
Now, it can be argued that the Treasury can issue all its bonds as 3m bills. In my opinion, it is a disaster in liability management.
Ramanan,
Thanks for your thoughts. I sent you a message at your blog to clear things up as well as to ask further question in a more organized fashion.
Thanks!
JKH- this was helpful for me. Thanks!
Hey all, the above commenting has been really useful for me and I'm glad this discussion is being had. But I saw most of the discussion regarding style vs. substance / factuals vs. counterfactuals only on the spending side. What is MMTs answers on the tax side? Taxation is entirely political, so if taxation was ever needed perhaps due to a past 'overheating' by the government that barrier seems to me very real and very un-ignorable. It calls into question the real operational ability for the 'government' to raise taxes when needed economically.
"It calls into question the real operational ability for the 'government' to raise taxes when needed economically."
Make it automatic for the people. :o)
"So how can we boost demand when we need it – and cut back when we don’t need it any more?
The solution: Give people more money when times are tough, and cut back when times are good.
The best way to do this is through a payroll tax holiday. When times are tough, people get laid off. But if your buddy gets laid off, you’ll get a raise!"
http://traderscrucible.com/2011/11/07/4-ways-to-change-the-fed-a-users-manual/
"Automatic for the people"
http://dixiedining.wordpress.com/2010/06/10/weaver-ds-is-still-automatic-for-the-people/
JKH: "Steve Keen, who's invented his own accounting along those lines"
I wonder if you could elaborate a bit. Any links? I also have been trying to bring together my thinking on MMT and Keen's work. Thx.
Steve is a great speaker and lecturer, but an accountant he ain't.
His models of double entry have one entry in them which mean that it just isn't a double entry bookkeeping model.
And yet it is fairly simple to alter the model so it is consistent with double entry bookkeeping and from that you can derive the standard MMT horizontal model with a simple write out.
I've got a post cooking about this point. I've been fighting Keen's QED tool this weekend and the double entry consistent models produce the same output as Keen's own models.
And what's nice about it is that the updated model diagram shows the asset side and the liability side circulating in their own right and growth is controlled by extending the bank's capacity to lend via a straightforward write back.
Ah understood, and fascinated to see what you come up with.
But: If the results are the same, why the big deal about double-entry?
Why I ask: I've run multiple multimillion-dollar businesses, with awesomely good cost- and income reporting -- the kind that MBAs dream of at night -- using Quicken -- a $29 single-entry bookkeeping package. Not QuickBooks. Careful tagging and categorization of transactions delivers all the reporting functionality that's normally dependent on the fairly byzantine account structure of double entry.
Doesn't double-entry exist basically to catch the kind of bookkeeping errors that humans make, but computers don't?
Related: Does MMT/Keen-style accounting involve/require accruals? Or is it basically cash accounting?
If accrual, that may provide a (partial?) answer to my question.
Is the advantage of double entry the additional outputs you mention -- graphs of assets and liabilities? Is DE necessary to get those?
Also, don't know what you mean by a "write out."
Thanks for indulging my curiosity...
"But: If the results are the same, why the big deal about double-entry?"
Double entry helps catch mistakes. Although the models currently produce the same output, they won't with more advanced models. I'm able to correct a couple of mild errors and simplify the model slightly conceptually due to double entry.
By sticking with double entry there is an inbuilt check that that catches basic errors. And it allows the models to be reviewed by accountants and other people familiar with the way banks work to make sure that the model is consistent and accurate against reality.
One feature of the double entry model I've come up with is that liabilities only move to other liabilities and assets only move to other assets once the initial money creation has happened. You may get linked transactions that change both sides at the same time, but you seem to be able to code the journals so that the liability side and asset side operate almost independently. It seems similar to an induction circuit in electronics.
A write out is when you move a value from an asset to a liability reducing both values towards zero and causing an amount of both to disappear (destruction if you prefer). A write back is the opposite when you move a value from a liability to an asset increasing both values away from zero - which is the creation process.
The double entry model is accrual based - one of the changes I make is that I accrue interest to the bank when it is due rather than forcing the bank to wait until it's been paid. This is in keeping with normal income recognition procedures and means it can be spent earlier by the bank, which may or may not affect the dynamics in more complex models.
Steve,
Try getting to an interpretation of the Steve Keen via this:
http://pragcap.com/the-deteriorating-macro-picture/comment-page-1#comment-25563
Neil and JKH (and Tom Hickey): Thanks! Truth told, more than I want to know. But your explanation, Neil, combined with scanning through the rest, is quite satisfying for one who wants to remain a (very curious) spectator, not a practitioner.
So my question for Steve would be: why *not* go double entry? (Before posting, btw, I sent him a note pointing to this thread.)
JKH,
Is there anything more recent than that? I'm pretty certain thinking on both side has evolved since then.
Perhaps it's time we organised a rematch :)
:)
Actually, I've been thinking of asking somebody for some Steve Keen links of my own, because I haven't been following him much since then!
Not that I'm disinterested in what he's doing; just haven't done so for some reason.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home