Overdrafts with 100% reserve banking

I was reading Frosti Sigurjonsson's proposal for monetary reform in Iceland. [Click on the document thingy halfway down the page.] His proposal is a variant on 100% reserve banking. I got stuck on page 72 and footnote 66, where he discusses overdrafts.

This post is just my attempt to get my own head clear on some of the questions this raises. My head is not 100% clear yet.

(Iceland's financial crisis was much worse than other countries', which concentrates their minds to consider radical reforms. But it would be wise for all countries, and that includes Canada, to keep at least one eye on the possibility of radical reforms. We don't really know why Canada has been relatively "lucky" so far.)

Suppose we stop keeping our money in our pockets. Instead we each have a box with our name on it at the Bank of Canada, where we keep our paper banknotes issued by the Bank of Canada. When I buy something from you, I send an email to the Bank of Canada, cc'ing you, instructing the Bank of Canada to take a $20 banknote out of my box and put it in your box. The Bank of Canada sends a reply-all email confirming it has done this. Then the boxes and paper banknotes all get destroyed in a fire. But it doesn't matter because the Bank of Canada has a computer record of how many notes are in each person's box, so everything carries on just as before. Paper money that we keep in our pockets is functionally equivalent to electronic money, except for convenience, muggers, etc.

100% reserve banking is exactly like that (all money is central bank money), except that there is limited decentralisation so that commercial banks manage some of the record-keeping and communications. Most proposals for 100% reserve banking require each commercial bank to keep a record of how much is in each customer's box, and the central bank to keep records of the totals for each commercial bank (with an occasional audit to check that the two sets of records match up). Frosti's proposal requires the central bank to keep records of how much is in each individual's box too (the commercial banks just handle the communication).

What happens when we introduce overdrafts into an economy with 100% reserve banking?

There are green notes worth +$1 each and red notes worth -$1 each. If you buy something for $20, you either have 20 green notes taken out of your box or 20 red notes put into your box. There is a limit to how many red notes you may have in your box, to prevent you buying unlimited amounts of goods, and accumulating an unlimited number of red notes. If you have only red notes in your box we call it an "overdraft".

Introducing red notes raises some interesting questions:

1. Who decides the limit on the number of red notes each individual may have in his box? Can it be decentralised so that commercial banks set those limits? If it is decentralised, is the commercial bank liable if the individual holding red notes becomes insolvent? If no, then commercial banks face moral hazard in setting that limit. If yes, then commercial banks can go bust despite 100% reserves, unless the commercial banks are required to hold green notes in their own boxes equal to their customers' holdings of red notes.

2. If the central bank: creates one green note plus one red note on demand; and destroys one green note plus one red note on demand, then the value of one red note will always be minus the value of one green note. (For the same reason that a $20 note will always be worth two $10 notes if the Bank of Canada swaps one $20 note for two $10 notes, or vice versa, on demand.) But then what particular monetary aggregate does the central bank control? Is it the the net stock of notes [the number of green notes minus the number of red notes]? (That is the direct equivalent of the Bank of Canada controlling [10 times the number of $10 notes plus 20 times the number of $20 notes].)

3. If the central bank controls the net stock of notes [green notes minus red notes], and commercial banks control the gross stock of notes [green notes plus red notes], then commercial banks do control the stock of money in one important sense, despite 100% reserve banking. Because red notes are media of exchange too. We can imagine an economy where the net stock of notes is zero, but the gross stock of notes is unlimited. (The simple New Keynesian model is like this.)

And that's as far as I've got.

Comments

[Jussi]

If banking sector is liable and required to hold green notes against overdrafts (1. in the post):

Then overdrafts, I think, require banking sector as a whole to keep more equity (green notes). If we assume that the equity is expensive (and I think in the model equity-green-notes sit in banks' boxes at the Central Bank) then banks' profitability limits their ability to run overdrafts. Is it then fair to say that the Central Bank controls indirectly the gross stock of notes too? This is quite alike the system we already have, the banks can expands money aggregates if their profitability allows it but only the Central Bank can easy it further. In that sense the proposal might not be that radical (didn't read it)?

Btw. I love these green/red posts/analogies, they are always insightful.

[rsj]

Cochrane also proposed a variant of this, except he argued that deposits be required to be backed by reserves or treasuries. Any quantity limit suffers from the same problem, namely it loses control of the interest rate and creates bank runs.

Either banks do not make loans in such a scheme (full 100% reserve banking) or they are allowed to make a few loans but are subject to bank runs should depositors withdraw their funds (so they fail and we are back to the no-loan scenario).

But there are many borrowers who cannot access the commercial funding markets except at extremely high rates, which are dominated by highly volatile risk premia. In such a world, the central bank can't control interest rates at all.

But it has been decided that control over interest rates, and the corresponding ability of individual households to borrow at low rates is more important than any philosophical benefit accrued from X% reserve banking.

It's interesting that in Cochrane's speech outlining his proposal, as soon as someone asked "Well, what if the demand for deposits exceeds the quantity of available government bonds", Cochrane immediately walked it back and started talking about risk taxes applied to banks -- effectively the current system, rather than any type of quantity restrictions.

Quantity rationing of the financial sector is always too rigid. Price rationing (e.g. interest rates, capital requirements, etc) are the only workable option, regardless of the appeal of quantity rationing to those who like to think in terms of quantities rather than prices.

Reserve banks were created as institutions to allow banks to avoid quantity rationing in times of crisis. Only after this happened was it possible for reserve banks to set policy interest rates and target macro-economic stability once the problem of bank stability had been solved.

If we were to try to continue targeting macro-economic stability while abandoning bank stability, then the whole system would collapse. If we ban banks from making loans, then lending will occur with some new institutions, called Nanks, and then we'd need to invent a new Reserve Nank to bail out the Nanks in their time of crisis, and we are back to the current system. Quantity rationing only works in a rational expectations world.

[Nick Rowe]

Jussi: thanks. I think that what you say is correct.

rsj: you lost me. Consider a world with only green notes. What would a "run" look like?

[Ralph Musgrave]

Nick,

Contrary to the suggestions in the 2nd half of your post, overdrafts or “loans” having nothing to do with the central bank - at least under the full reserve system advocated by Milton Friedman and Laurence Kotlikoff. And more or less the same goes for Positive Money’s version.

Under full reserve, loans are made by entities or “banks” or bank departments which are funded just by shares or largely by shares. Thus contrary to your suggestion that “commercial banks can go bust despite 100% reserves”, those lending banks (or bank departments) cannot go insolvent: it’s near impossible for a bank funded just by shares to go insolvent. And that makes the banking system much more stable.

I ASSUME the Iceland model is similar to Friedman, Kotlikoff and Positive Money’s, and if so, that renders your final points “2” and “3” obsolete.

For Friedman, see Ch3 of his book “A program for monetary stability” under the heading “How 100% reserves would work”. As to Kotlikoff, Matthew Kline does a good introduction to K’s system here:

http://www.bloombergview.com/articles/2013-03-27/the-best-way-to-save-banking-is-to-kill-it

There’s also a good paper on this subject by Adam Levitin:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2532703

[Ralph Musgrave]

Rjs,

You are right: Cochrane proposed a variation on this system. See here:

http://www.hoover.org/news/daily-report/150171

The rules of full reserve are actually being imposed on money market mutual funds in the US. Good thing too in my view. See:

http://www.euromoney.com/Article/3426686/Money-markets-Short-term-bank-funding-costs-hit-by-Fidelity-move.html

Re your claim that under full reserve (FR) the central bank loses control of interest rates, I’m baffled. What’s to stop it printing base money and buying up government debt just like CBs do at present? Nothing that I can see. Or what’s to stop it wading into the market and offering to borrow at above the going rate for risk free loans, if the CB wants to raise interest rates? Nothing in principle.

Moreover, even if FR does reduce the CB’s ability to adjust interest rates, the reaction of one group of FR enthusiasts (Positive Money and Richard Werner) would be “we couldn’t care less”. Reason is that they specifically argue for fiscal measures rather than interest rate adjustments when trying to control AD. See:

http://www.positivemoney.org.uk/wp-content/uploads/2010/11/NEF-Southampton-Positive-Money-ICB-Submission.pdf

Re your claim that FR “creates bank runs”, I’m baffled. As Cochrane specifically explains, when a bank is funded just by shares, runs just don’t happen.

Your 2nd para says “Either banks do not make loans in such a scheme (full 100% reserve banking) or they are allowed to make a few loans but are subject to bank runs should depositors withdraw their funds (so they fail and we are back to the no-loan scenario).” Nope. Depositors DO NOT fund loans under FR: its shareholders that do the funding. And shareholders can't/don't run.

[Jussi]

I think in the model* the bank equity needs to match the amount of overdrafts. So there is no room for banks to expand effective gross stock of notes (effective as excluding green-notes-equity) without the Central Bank offering more liquidity.

*If banking sector is liable and required to hold green notes against overdrafts

[Nick Rowe]

Ralph: loans and overdrafts are different. I can give you a loan (I give you some of my green notes and you give me an IOU). If the bank gives you a $100 loan, in a 100% reserve system, the bank lends you 100 green notes, puts them in your box, and you give the bank an IOU. With an overdraft, you only owe what you actually spend. So if the bank gives you an overdraft limit of $100, but you only spend $20, you only have 20 red notes in your box, not 100.

Jussi: I think it depends on whether we count those green notes in bank's equity box as money.

[Jussi]

@Ralph

"Cochrane also proposed a variant of this, except he argued that deposits be required to be backed by reserves or treasuries"

Isn't this just the heart of the 100% reserve banking? The additional/initial money/reserves/treasuries need to come from somewhere - the private part of the economy cannot expand money by itself at all (only velocity can change). So the government/Central Bank needs to limit or accommodate the expansion. How the proposals tackle this? What rules or who decides how much credit can be expanded in the proposals?

[Jussi]

@Nick: exactly, that's why I tried to introduce effective gross stock of notes. I think in practical terms we shouldn't as green notes in the equity box don't circulate anymore (they are kind of back in the Central Bank again).

[Roger Sparks]

"There is a limit to how many red notes ..........."

Try adding to your thinking that the limit is based on green notes. More exactly, any individual who holds green notes can exchange his green notes for red notes. His does this by emailing the bank (taking advantage of a bank service) instructing the bank to deduct green notes from his account in in favor of friend XXXXX with the stipulation that friend XXXXX will repay the green notes in the future at YYYYY date. Friend XXXXX signs a note payable to the bank acknowledging that he is accepting his friends generosity.

With this limitation, red notes are not overdrafts. Red notes are a link to green notes used by another player.

The exchange of green notes for red notes is a bank service. The bank may charge for the service.

If friend XXXXX fails to repay the green notes at future YYYYY date, something may happen (or not) to the linked red notes. The 'something' in case of loan failure is the next step in the green-red story.

Thus ends this suggestion for continued thinking on the green-red money analogy.

[Ralph Musgrave]

Hi Jussi,

Re “who decides how much credit can be expanded in the proposals” Positive Money and Richard Werner propose that the amount of additional base money spent into the economy in the next six months or whatever is decided by some sort of committee very much like the existing central bank committees that decide on interest rate adjustments and QE. As to the EXACT WAY that money is spent (which of course is a POLITICAL decision) that’s left to voters and politicians. See bottom of p.10-12 here:

http://www.positivemoney.org.uk/wp-content/uploads/2010/11/NEF-Southampton-Positive-Money-ICB-Submission.pdf

In short, STIMULUS is decided by the above sort of committee just as it is at present.

As to “credit” in the “lending” sense of the word, that’s done by banks or bank departments which are funded by shares or funded largely by shares. Thus the total amount loaned is determined very much by market forces: how much shareholder money can banks attract and at what price versus what demand is there for borrowed funds and at what price.

Milton Friedman advocated the above sort of “create base money and spend it into the economy in whatever amounts are needed to maintain full employment” in a 1948 paper. See para starting “Under the proposal…” here:

http://0055d26.netsolhost.com/friedman/pdfs/aea/AEA-AER.06.01.1948.pdf

Though far as I know he didn’t advocate that SPECIFICALLY in connection with his full reserve / 100% reserve proposals.

[Mike Sproul]

Nick:

Overdraft=loan

So you are asking what happens if a 100% reserve bank is not a 100% reserve bank.
And the number of red notes I can have in my box is limited by the amount of collateral I can give to the bank.

[Nick Rowe]

Jussi:

1. The commercial bank goes to the central bank and demands 100 green notes plus 100 red notes be put in its box. The central bank does this. (This is the equivalent to a commercial bank asking the central bank to swap one $20 for two $10 notes.)

2. The commercial bank then tells customer A he is allowed to have an overdraft of up to $100.

3. customer A spends $60 buying a bike from customer B. The bank transfers 60 red notes from its own box into customer A's box, enters "$60" on customer A's blank signed IOU to the bank, and transfers 60 green notes from its own box to customer B's box.

4. So customer As box has 60 red notes, customer B's box has 60 green notes, and the bank's box has 40 green and 40 red notes.

The net money supply stays exactly the same throughout. But the sum of green+red notes in customers' boxes has increased by 60+60=120.

[Jussi]

@Ralph: thank you for explanation and pointers. Will read.

I see that there is the committee and upper limit (if/when the reserve ratio is binding) of private credit expansion already in the current banking system so it wouldn't be that different. But the proposed system feels stricter as the money (reserve) multiplier is one and thus all the credit given needs to be met by additional reserves. But this is only a matter of degree and can be addressed.

I can see how spending in the existence vs. additional reserves against collateral/purchases might have a difference in terms of aggregate demand (at least in practice, I guess MM theory and Nick disagree) and thus employment.

This is probably already off-topic but would you say that pros of the proposals would be less/no moral hazards (no deposit guarantee), no bank runs and likely less instability? Cons might be at least slightly less growth as savings/investments are not subsidized anymore?

[Majromax]

I think you can see the plan's outline most clearly on page 82/85, "Availability of bank credit following the switchover."

The main thrust of the plan is to demonetize savings. Funds made available for loan would be put into the bank's "Investment Pool Account," which is an aggregate of individual "Investment Accounts" (page 71/74), which are most emphatically not demand deposits. Demand deposits will be secure but not interest-bearing; investment accounts are on the other hand not guaranteed.

In terms of managing the money supply, the proposal permits the central bank to make loans to commercial banks to increase the supply of base money. However, it does a far longer job of selling MMT-like "money creation for public use" approaches.

In practice, I don't see how well this will work. Either the system would be mananged for interest-rate stability (that is, explicitly reducing the risk of investment accounts), or it will lead to serious interest-rate volatility. If the latter happens, I don't see any reason that more of the Icelandic economy would become euroized or dollarized.

[djb]

"all money is central bank money"

all loaned money anyway

because is the government paid me money for services rendered and i put it in the back, they couldnt touch it to lend it out

with 100% reserve banking

but then again i guess i would have pay a portion to the bank for their services

ie interest rate would have to always be negative

an overdraft would have to be a loan and therefore come from 'central bank money"

[Too Much Fed]

Mike Sproul said: "Overdraft=loan"

I agree that overdraft = "red money" = loan/bond. Out in the real world, there is only "green money" and bonds.

[Ralph Musgrave]

Jussi,

“I guess MM theory and Nick disagree”. I don’t see a big clash between MMT and full reserve. In fact there was an article on the Positive Money site a few days ago considering whether MMTers and the Pos Money lot could work together. MMTers (particularly Roger Erickson on Mike Norman’s site) tend to favour the idea of simply “creating fiat” (as they put it) and spending it in a recession.

Re your last para, there is less moral hazard in that deposits (or more broadly “stakes in banks”) which fund risky loans are no longer guaranteed by taxpayers. The only guarantee is for deposits where the relevant money is lodged in a near totally safe manner (e.g. at the central bank). So that guarantee should cost taxpayers little or nothing. Ergo there’s no subsidy of the banking system there.

Re your last sentence, you suggest that because money deposited at banks which is loaned on to mortgagors and businesses is no longer backed by taxpayers, that there’ll be less investment. Correct. But I don’t agree that that reduces growth. Reason is that ANY SUBSIDY (including investment subsidies) missallocate resources, unless a good reason is found for the subsidy (e.g. market failure or some sort of social reason). So… the initial effect of introducing full reserve is to cut investment and hence AD, but that’s easily compensated for by creating and spending new base money into the economy. The net effect is less loan / investment based activity and more non-loan and non-investment based activity.

So the net effect (in view of the removal of an unjustified subsidy) should be to INCREASE GDP, not reduce it.

[djb]

also if every time the fed issues a credit "green" so the bank can loan then they would have hold a red at the fed

the banks could then send this money to the lendee since its "central bank money"

not deposit money

fed not being subject to 100% reserve

[Jussi]

@TMF: "Out in the real world, there is only "green money" and bonds."

What is the color of money on your credit (overdraft) card?

@Ralph:

Sorry the confusion and use of an acronym, my MM stood for "Market Monetarism".

Yes, I agree with the misallocation part - thinking it more the reduced instability might also encourage more investments. But I'm not sure how GDP is necessarily up because it depends on which effect dominates. But I think in utility terms we would be better off as there would be less distortion towards non-optimal savings - I mean max growth comes by maximizing investments but that is hardly the best world to live in.

[csissoko]

I think your analysis in point 1. is entirely accurate. My reading of footnote 72, however, indicates that any bank that authorizes an overdraft must indeed hold a green note (presumably in its operational account, p. 75) for every red note in a customer account. (That is, I think what Jussi is calling equity would under the proposal actually have to be held in the operational account.) I think this is what the authors mean when they write "When a customer with an approved overdraft draws down the overdraft, he is borrowing from pre-existing sovereign money owned by the bank." These, after all, must be green notes, right?

You then posit in point 2 that the central bank stands ready to hand paired red/green notes on demand. But I think this presumption is precisely what the central bank is unwilling to do in a narrow banking framework -- because it gives control of the money supply to the banks, which can then issue unlimited overdrafts. (Which roughly speaking is I think what Woodford's "cashless" model seeks to do.) I think the idea is that the central bank controls the stock of green notes, and banks can create their own red notes to offer people, but for every red note created by a bank there must be a green note sitting in the bank's reserve account at the central bank. In short, the bank can transfer control over the value of the green notes that it owns to customers, but cannot create money.

The footnote continues "From the customer’s point of view, the experience of using overdraft in the sovereign money system will be very similar to using an overdraft in the current system." Which I think is only superficially correct. (I think your model is reading more into this statement than the authors of the proposal intended.) Functionally overdrafts will be similar. In practice, however, requiring banks to fund overdrafts will prevent them from creating money and undoubtedly have a significant effect on the availability of overdrafts.

On a related note, I try to explain simply in two posts why new monetarist models indicate that narrow banking proposals that eliminate expansion of money by banks through overdrafts are a bad idea.
https://syntheticassets.wordpress.com/2015/03/28/new-monetarism-and-narrow-banking/
https://syntheticassets.wordpress.com/2015/03/31/new-monetarism-and-narrow-banking-take-two/

[Too Much Fed]

Jussi said: "What is the color of money on your credit (overdraft) card?"

The color of the demand deposits that are borrowed is green.

The credit card itself is instructions to issue a "green" bond to the credit card issuer. The bond issued to the credit card issuer is not "money".

[rsj]

Nick,

'you lost me. Consider a world with only green notes. What would a "run" look like?'

A bank run is the failure of the bank to be able to borrow green notes against good collateral.

Banks on the one hand lend out green notes, and on the other hand they borrow green notes. A depositor depositing green notes is actually lending the notes to the bank. But banks do not need to be passive and sit around looking pretty hoping that a depositor will show up before they make a loan. They make loans to borrowers who are qualified and the bank itself goes to commercial (and interbank) funding markets to *borrow* the green notes that the bank needs to meet its regulatory requirement to hold green notes as well as to meet any actual green note outflows.

But the green notes that the bank borrows are shorter term than the duration of the loans (e.g. depositors can withdraw their money any time, and the bank borrows from short term funding markets). This is because the bank really only needs to borrow green notes on a short term basis, to meet fluctuations between outflows and inflows.

A run on the bank would be a situation in which the bank cannot borrow enough green notes to meet its regulatory requirement to hold notes or its actual requirement to provide notes to those who withdraw. This could happen because the short term market does not trust the bank while at the same time depositors are withdrawing. I.e. the lenders to the bank have lost confidence in that bank. In this type of situation, the bank needs to be able to turn to the Reserve Bank and be able to get whatever quantity of green notes it needs. And if the lenders to the bank know that the bank can get whatever quantity of green notes it needs, whenever it needs them, then they will not lose confidence in the bank as long as the collateral is sound. That prevents bank runs, but at a cost of the central bank losing control (or not having any credibility) over the *quantity* of reserves outstanding. It cannot tell a bank -- no more green notes for you! -- because then the bank is liable to bank runs again.

[Nick Rowe]

csissoko: "These, after all, must be green notes, right?"

Yes, I think so.

"(Which roughly speaking is I think what Woodford's "cashless" model seeks to do.)"

I think so too. I read Woodford's "cashless" model as having an equal number of green and red notes (so net money is zero), but an unlimited number of green+red notes. (The modeller enforces the no-Ponzi condition by assumption).

" I think the idea is that the central bank controls the stock of green notes, and banks can create their own red notes to offer people, but for every red note created by a bank there must be a green note sitting in the bank's reserve account at the central bank."

Maybe. But then red notes created by the Bank of Montreal and sitting in my bank account at BMO would be assets of BMO, and if BMO also has one Bank of Canada green note for every red note it creates, that seems like a 200% asset/liability ratio. (I'm not sure I'm right on that.)

[csissoko]

Re Woodord's "cashless" model. Not surprised you think so too. I'm pretty sure I got that understanding from you! Glad I didn't mess it up.

I think the accounting for this is a little complicated. Once a red note has been issued by the bank, the corresponding green note becomes something similar to an encumbered asset. Probably collateral-type accounting would be needed, where it is necessary to note whether the asset that you nominally own has been pledged as collateral to someone else.

[csissoko]

I think one of two accounting frameworks would have to be used:

(i) Bank holdings of green notes in a central bank account are counted as assets only on a net basis, that is, after subtracting off the number of outstanding red notes issued by a bank.

Or

(ii) Bank holdings of green notes in a central bank account are counted as assets on a gross basis. However, for every red note issued by the bank, three new entries are created:
Liability to recipient of overdraft
Asset due from recipient of overdraft
Liability to central bank (borrowing fully collateralized by green notes)

[Jussi]

@TMF:

The credit card itself is instructions to issue a "green" bond to the credit card issuer.

The red money itself is instructions to issue a "green" bond to the red money issuer?

[rsj]

Ralph,

"The rules of full reserve are actually being imposed on money market mutual funds in the US."

Huh? I can write checks against my stock holdings in brokerage account at this very moment. What you are talking about just adds more costs and fees, it doesn't actually prevent banks from providing money-like services held against less liquid assets, as that's kinda the whole point of banking.

"Re your claim that under full reserve (FR) the central bank loses control of interest rates, I’m baffled. What’s to stop it printing base money and buying up government debt just like CBs do at present? "

That is not my claim. My claim is that if the CB tries to fix the *quantity* of reserves, it loses control of interest rates. Of course when you say that the CB can buy up X, you mean that it must abandon it's previous quantity peg.

And in reality, the CB cannot control the quantity of reserves because banks need a certain amount of reserves, and if the CB does not make this amount available, then there will be a banking crisis. So in practice, the CB has a choice in setting the quantity to be greater than the minimum that banks need. But if the quantity is greater than what banks need, the rate drops to zero. So the CB, if it is to control rates, needs to set the quantity to be exactly what banks need. It doesn't really have control over quantity.

Now it can choose to pay interest on reserves, but in this case there is no difference between CB and Treasury liabilities.

[Max]

This is the kind of thing that gives full reserve banking a bad name. The paper proposes not to pay interest on deposits. So deposits will be "taxed" at some random fluctuating rate depending on prevailing interest rates. That's dumb. (Yeah, that's how currency works, but there's an excuse: it's tricky to pay interest on currency).

[Frank Restly]

Nick,

"If the central bank controls the net stock of notes [green notes minus red notes], and commercial banks control the gross stock of notes [green notes plus red notes], then commercial banks do control the stock of money in one important sense, despite 100% reserve banking. Because red notes are media of exchange too. We can imagine an economy where the net stock of notes is zero, but the gross stock of notes is unlimited."

I buy a widget for $20 green notes. Do I get the choice of either paying $20 green notes or accepting $20 red notes even if I have the $20 green notes?

This is not an obvious choice. My liquidity is improved if I accept the red notes assuming they are as liquid as the green notes. Meaning I later have an easier time selling a whatsit that I produce since my buyer does not need green notes to buy it from me, he just accepts the whatsit and the $20 red note that I give him.

The central bank may not want to give the commercial bank additional red notes if there are already too many in circulation relative to green notes.

And so the commercial bank trying to keep me as a customer deducts $19.95 in green notes from my account instead of crediting me with $20 in red notes.

The liquidity premium is born.

[Nick Rowe]

csissko: "I think the accounting for this is a little complicated. Once a red note has been issued by the bank, the corresponding green note becomes something similar to an encumbered asset."

That sounds right to me.

If those are the rules needed to ensure that a bank that allows overdrafts can never go bust (which is the whole point of 100% reserve banking), I'm beginning to wonder if such a bank would ever allow its customers to have overdrafts. The customer seeking an overdraft with a $100 limit would simply borrow $100 instead, and put 100 green notes in his box. The non-payments half of the bank (that operates just like a non-bank financial intermediary) could just as easily lend the customer those 100 green notes.

I guess that the point of this post is that there's some sort of contradiction between overdrafts and 100% reserve banking. Wish my head was clearer.

[Ralph Musgrave]

Max,

Not paying interest on deposits is not a huge change from the EXISTING system. E.g. I pay about £12/month fees to my bank for my current / checking account and get no interest. If I shopped around, I could probably find a better deal, but I can’t be bothered.

Second, whether interest is paid on deposits under full reserve banking depends on whether (as proposed by Milton Friedman) deposited money is invested in government debt (where it obviously earns interest) or whether deposited money can only be kept in the form of base money (which normally pays no interest). I favour the Friedman option.

Third, the ability of banks to pay interest on deposits under the existing system relies on a fraud, or on taxpayer backing for private banks. That is, there is no such thing as a totally safe set of loans or investments (made by a bank or anyone else). Thus the promise by banks to return $X to customers for every $X deposited AND PAY interest is a flagrant self-contradiction: it’s a fraudulent promise. It’s a promise that can only be made good 100% of the time if taxpayers stand behind private banks.

Under full reserve, depositors have to get real. If they want total safety, then their money is kept in a more or less totally safe manner, but that ipso facto means their money is not loaned out (except possibly to government) so the money earns little or no interest.

Alternatively, depositors can have their money loaned on, which means that interest or more interest is earned, but in that case depositors take a hair cut if the loans go wrong. That means that depositors effectively become shareholders: they are no longer depositors.

[Ralph Musgrave]

Rsj,

Re money market mutual funds, the change in the rules governing MMMFs don’t come into force for some time yet (I think it’s about 18 months time). So that explains why you can write cheques against a fund which invests in corporate bonds at the moment.

And even after the rules ARE CHANGED, it would be possible to have a system where people can write cheques against the above sort of fund without breaking the basic rules of the game. The basic rule is that only funds which invest in short term government debt or just boring old base money can promise to return $X to customers for every $X deposited. In contrast, funds which invest in anything more risky have to let the value of customers’ stakes in the fund vary with changes in the value of the underlying assets.

Whether it’s desirable to actually let customers write checks against the latter sort of fund, I’m not sure: the pros and cons are a bit complicated, seems to me.

Re interest rate adjustments under full reserve, I think you make some good points. I also think this issue is too complicated to be fully addressed in comments after a blog post. But I’ll throw in just one point, as follows.

I agree with the claim often made by MMTers that if the state issues more liabilities (“private sector net financial assets” (PSNFA)) than the private sector wants to hold, the state will have to pay interest to those holders to induce them to “hold”. Thus the state CAN push up interest rates by expanding the stock of PSNFA, and then paying interest on the latter. Put another way, governments can borrow, but doing so will tend to raise interest rates. Having done that, the state can then cut that interest rate by printing money and buying back some of that govt borrowing.

[Oliver]

" I think the idea is that the central bank controls the stock of green notes, and banks can create their own red notes to offer people, but for every red note created by a bank there must be a green note sitting in the bank's reserve account at the central bank."
Maybe. But then red notes created by the Bank of Montreal and sitting in my bank account at BMO would be assets of BMO, and if BMO also has one Bank of Canada green note for every red note it creates, that seems like a 200% asset/liability ratio. (I'm not sure I'm right on that.)
I'm pretty certain that's not right.
Firstly, notes created by the BMO are its liabilities, not its assets. So, in a 100% reserve environment, every green note created by a commercial bank has a corresponding red note created by the CB sitting on the opposite ledger on the commercial bank's books. Secondly, the asset / liability ratio for every institution, including the CB, is always 1:1, per identity. Now, if one increases the requirement for one type of asset (reserves), one needs to first know how they get onto bank books and then one can decide what that might mean for the ability of the bank to make new loans.

From Wikipedia: A central bank may introduce new money (read = reserves) into the economy by purchasing financial assets or lending money to financial institutions.

Let's assume that purchases are the norm (that's true for the US, as far as I know). Bank A has made new loans during a period and created new red notes / demand deposits (which we'll assume remain on its books) in the process. It now needs new green notes / reserves to cover the increase in red notes / demand deposits. Assuming it wants to keep banks from going bust / not force them into borrowing reserves, the central bank will purchase (or sell) other bank held assets, most notably government bonds, and exchange them for reserves at the end of a maintenance period in the exact amount required to keep the interbank lending rate of reserves on target. It can do this for 0%, 10% or 100% reserve requirements. But since reserves do not normally pay interest (as opposed to other assets such as government bonds) a higher reserve requirement makes deposit banking more expensive. Banks can circumvent this by swapping red notes for other types of liabilities, say time deposits (which tend to be more expensive than demand deposits, though). Central banks can also pay interest on its reserves. It also means there is a natural market for the types of securities that central banks buy / sell.

This is a simplified description of my understanding of a typical modern monetary system. But, even assuming I'm right, that of course does not mean it could't be otherwise. What the 100% reserve folk seem to be getting at is that their proposed system would keep banks from making new loans above the amount of reserves already in the system. Assuming it is the expansion and not contraction of loans / deposits that they wish to control, that must be true for any non zero reserve requirement. If I limit the amount of reserves to $1'000 at 10& reserve requirement, I have de facto limited the maximum amount of credit money to $10'000. And even at 0%, the amount of new deposits could just be limited by decree.

But such limits could only be achieved if central banks closed their discount windows and gave up targeting lending rates. Apart from the fact that it is those two things that central banks were invented for, it is also my opinion that that would not actually work. A world with many different commercial banks that emit means of payment at par with the official medium of account but with no LOLR and interest rate targeting central bank is a nonsensical counterfactual. Interest rates would move so eratically that banks would either suffer constant runs or be forced to give up their peg to the official medium of account. It would be like the euro system on speed. I claim this is not an artefact of modern money, but an inherent trait of money as such. Homgenising the value of money across heterogenous agents requires giving up quantity targets.

[Jussi]

I think I got it but now I'm not sure.

What is the actual distinction of green and red notes? Is it only about the hierarchy, I mean CB liabilities are green for banks and bank IOUs (deposits) are green for customers but red for the banks. Having 100% reserve banking means IMO then that the bottom red notes (e.g. bank deposits) have one-to-one correspondence to the top green notes (reserves) and they have the same (CB) credit qualities - the amounts are the same. But if that is true, I'm not sure why not circulating only green notes in the first place? What is the advantage to create a red note if there needs to be a corresponding green note locked up on an upper (operational?) account?

So is the proposed system the same than using central bank money (or accounts) only?

Did I miss something big?

[David Andolfatto]

Nick,

I haven't had time to read that report or even the comments above, but I'll go ahead and ask anyway: Why would we want to permit overdrafts in the first place in a 100% reserve banking system? If people need to borrow money, there'd be a separate credit market on which they could do so. That's the idea behind separating money from credit.

David

[Nick Rowe]

David: I'm slowly coming to the same conclusion as you. The Report just discusses overdrafts in passing, but I thought that overdrafts seemed "problematic" in a 100% reserves system, so I wrote this post, trying to get my head around it.

[Frank Restly]

Nick,

Does 100% reserve banking keep an individual commercial bank from failing (assuming there are many commercial banks)? I don't think so. 100% reserve banking will not stop banks from competing with each other. At some point, a poorly run commercial bank will have an excess of red notes relative to green notes and could be considered insolvent / bankrupt.

I (as bank customer) would prefer to have red notes credited to me rather than green notes taken away when I purchase a good. Commercial bank would rather take green notes from me rather than issue red notes to me when I purchase a good. So, I shop around for commercial bank that has excess red notes they are willing to give me or for a commercial bank that will take fewer green notes from me when I purchase a good.

I (as bank customer) would prefer to have green notes credited to me rather than red notes taken away when I sell a good. Commercial bank would rather take red notes from me rather than issue green notes to me when I sell a good. So, I shop around for commercial bank that has excess green notes they are willing to give me or for a commercial bank that will take fewer red notes from me when I sell a good.

So there are really two liquidity premia - one for the seller of goods and one for the buyer of goods. Seller of goods prefers liquidity of receiving green notes, buyer of goods prefers liquidity of receiving red notes.

If there is only one bank (the central bank), then the liquidity premia should be exactly the same when the central bank maintains 100% reserves (equal green notes and red notes).

If there are multiple commercial banks all competing with each other for profits - then the liquidity premium for green notes shifts - they will be in higher demand by both sellers of goods AND commercial banks. And so green notes would trade at a premium to red notes even though the quantities are held to be the same by the central bank. If the difference in liquidity premia is sufficient - you could get a recession.

The central bank must be able to increase the quantity of green notes relative to red notes if it wants profit seeking commercial banks.

[Oliver]

If people need to borrow money, there'd be a separate credit market on which they could do so. That's the idea behind separating money from credit.

How does money initially enter the (non-bank) economy if not by somebody borrowing it? And isn't it the aim of 100% reserve banking to have better or even complete control over credit growth?

[Oliver]

I'll rephrase that. How does money enter the non-bank economy without first being credit money?

[Jussi]

"How does money enter the non-bank economy without first being credit money?"

By QE which buys assets from non-bank sector or fiscal stimulus?

[Jussi]

"At some point, a poorly run commercial bank will have an excess of red notes relative to green notes"

This is not possible by the definition/rules, the bank needs to match red ones with green ones - one way or the other.

[csissoko]

Nick: "The non-payments half of the bank (that operates just like a non-bank financial intermediary) could just as easily lend the customer those 100 green notes. I guess that the point of this post is that there's some sort of contradiction between overdrafts and 100% reserve banking."

I agree with you that overdrafts, even if permitted on a fully funded basis, are unlikely to take place in a 100% reserve banking framework.

I think though that the author's desire to allow for overdrafts gets to the heart of why 100% reserve banking proposals are in the end unconvincing. These proposals are premised on the idea that banks are "just" intermediaries between savers and lenders, so nothing will be lost by forcing debt out of the payments system.

To the degree that the payments system only functions well, because it facilitates transactions by making debt more available to account holders, and that it does so easily because banks don't need to source the funds from savers before they offer such credit to account holders, then the 100% reserve banking will reduce economic activity.

[Frank Restly]

Jussi,

Under Nick's #3, the commercial banking system as a whole can't have more green notes than red notes (or vice versa) because of central bank intervention.

That does not mean that commercial banks competing against each other can't individually have more of one and less of another.

[Jussi]

Frank,

I assumed standard 100% reserve banking where all the banks individually are required to keep the reserves at 100%.

[Frank Restly]

Jussi,

"I assumed standard 100% reserve banking where all the banks individually are required to keep the reserves at 100%."

Next to impossible to do. I (as a bank customer) can buy goods through one bank using red notes. That bank would then need to match my purchase with another that uses green notes to balance it's transactions.

If I am playing the liquidity game, I would keep my green notes at one bank and keep my red notes at another. When I sell something through my green bank, I would have no red notes to give up (preserving my liquidity) and when I buy something through my red bank, I would have no green notes to give up (again preserving my liquidity).

If enough people play this game, then there becomes two types of banks - all red banks used by consumers and all green banks used by producers. If profits / losses and the solvency of the bank is measured in net green notes versus red notes, then consumer banks (red banks) fail regularly while producer (green banks) live on.

Finally, if all commercial banks try to maintain 100% reserves, then the quantity of transactions may fall - I want to buy a $1 million airplane with red notes but my bank doesn't have a matching green note transaction to offset it with.

[Jussi]

Frank,

You cannot game it as customer if a bank is reserve constrained it will decline your loan - no more red notes and game over. Thus excess loans are not made; only the CB can expand credit - this might be good or bad.

Transactions are different though and won't fail, you can transact with red notes as you wish but banks will always mirror the transactions with corresponding green notes (think like they would be paired) - all banks will always be at 100% reserves.

[Frank Restly]

Jussi,

Suppose that at the end of each working day / week all of the commercial banks got together and tried to all balance their books by trading green notes and red notes. This situation would not be "real time" 100% reserves which is next to impossible, but rather fixed instant in time 100% reserves.

If all of the commercial banks are profit seeking and competing against each other, then a premium would develop for green notes versus red notes. The central bank could allow this premium to exist or could offset it by maintaining a floating reserve ratio.

[Frank Restly]

Jussi,

Yes I can game the system - I and a bunch of my friends move all of our red notes to one bank and all of our green notes to another bank. The quantity of red and green notes has not changed, but the distribution has.

That change in distribution could negatively affect the bank that has accepted our red notes and positively affect the bank that has accepted our green notes.

[Oliver]

Jussi:
By QE which buys assets from non-bank sector or fiscal stimulus?

No, because the CB can only buy in the secondary markets which means somebody else must have bought first. And if you start from scratch, that cannot happen with central bank money. My point: although CB money may be alpha in the sense that it is the bench mark, credit money is alpha in another sense, namely that it is logically prior. It is the original financial asset that precedes both private securities and CB money (in my admittedly not very humble opinion, that is).

[Jussi]

Oliver,

Do you assume the Central Bank can only buy government papers? The (Central) Bank can buy/warehouse almost anything against money (its liabilities), gold, food or labor if we go really primitive. And the government can always first spend.

Frank,

"Suppose that at the end of each working day / week all of the commercial banks got together and tried to all balance their books by trading green notes and red notes. This situation would not be "real time" 100% reserves which is next to impossible, but rather fixed instant in time 100% reserves."

I think the US banks did meet in New York quite like that before the Fed was created. And the Fed was partly created to make the system more stable in the case a bank couldn't attract deposit or other banks' credit. The money market (sorry Nick) and the Fed is/would be there to ensure that payments go smoothly - I wouldn't bet against the Central Bank.

[Max]

Ralph, if the government makes deposits onerous then people will seek workarounds. Example: money market funds were invented in the '70s to circumvent interest rate caps (risk-free interest rates were >10% but banks were prohibited from paying more than 5%).

Deposits can be very safe if nobody uses them. But then, whatever they are using instead will be the new focal point of a financial crisis.

So the trick is to make deposits safer without reducing the quantity of deposits.

[Michael Byrnes]

This system would be really rough on the colorblind.

[Niveditas98 .]

Full-reserve banking is an oxymoron. Full-reserve banking is not banking. It is a storage/payments service.

Credit intermediation and maturity transformation will be done by entities which will not be called "banks" because they won't take deposits, but will have exactly the same issues as existing non-full-reserve banks. We will then proceed to reinvent the entire structure of regulation that deposit-taking fractional-reserve banks are subject to, and end up with the same system that we have today.

[Ralph Musgrave]

Max,

There is plenty of truth in your claim that “if the government makes deposits onerous then people will seek workarounds.” I’d put it differently, and as follows.

Banks will always try to claim that money deposited with them is totally safe, at the same time as using that money in a way which is clearly not totally safe. Doing the latter means extra profits for banks and more interest for depositors, until the whole thing crashes and the taxpayer comes to the rescue. It’s win, win, win for banks and depositors, and lose, lose, lose for taxpayers.

Niveditas98,

I don’t agree that full reserve banks will involve “exactly the same issues as existing non-full-reserve banks”. It’s near impossible for a full reserve bank to fail, though given incompetent management, the value of its shares would decline which would make it a take-over target.

As to the safe half of the industry under FR, that cannot fail. As to the lending half, that’s funded just by shares under Kotlokoff and Friedman’s versions of FR, so that can’t go insolvent either. In contrast, the proposed Icelandic version, which is pretty much a copy of the Positive Money version, DOES INVOLVE the possibility of failure (as Positive Money admit).

However, those making deposits in the lending half sign up for the possibility of failure, thus there shouldn’t be any need for taxpayer funded rescues. However, agreeing to the possibility of losing all your money in effect makes you a shareholder, thus I don’t see the point of that aspect of the Positive Money version. I prefer Kotlikoff / Friedman.

[Roger Sparks]

"What happens when we introduce overdrafts into an economy with 100% reserve banking?

There are green notes worth +$1 each and red notes worth -$1 each. If you buy something for $20, you either have 20 green notes taken out of your box or 20 red notes put into your box."

At this point in your post, you introduced disharmony into your model. You no longer have 100% reserve banking, you have a two currency banking system.

If your goal is to continue with a 100% reserve banking model, you need to link the red notes to green notes, with a goal of the overdraft flowing to one box containing green notes. 100% reserves does not preclude lending, it assigns a loser (a box containing green notes) to every loan in case of loan failure.

[Oliver]

Jussi

Do you assume the Central Bank can only buy government papers? The (Central) Bank can buy/warehouse almost anything against money (its liabilities), gold, food or labor if we go really primitive. And the government can always first spend.

No, no need for that assumption. But I do assume that central banks do not bid on primary markets for whichever assets they buy. So food or labour are not on the menue. The central bank as the government's bank is a special case. But the fact remains that you don't get transfer payments or any other payment in cash or reserves. Bank money comes first and then you can exchange it for central bank money on demand. The creation of new means of payment ex nihilo (as opposed to swapping some existing asset for means of payment) is the product of a tripartite transaction between two private entities and a commercial bank. Credit where credit is due.

[Niveditas98 .]

Ralph, I don't mean that full-reserve "banks" will have failures -- I mean that other entities will spring up to perform the economic function that fractional-reserve banks currently do, and they will create the same issues. It is fantasy to believe that the failure of these entities will not cause problems because they are funded by "equity".

Fractional reserve banks when they originally arose did not have deposit insurance. Deposit insurance was created to mitigate the economic stresses caused by their failures, and it has worked quite well.

Fractional reserve banking serves a useful economic function, which is why it exists. I think of it as essentially providing insurance: after all, why do people want to hold their savings as deposits? It is because of uncertainty around when they will need money. Just as it is more efficient to pool resources in an insurance company to provide fire insurance, rather than each homeowner attempting to self-insure, it is more efficient for depositors to pool their deposits together and lend out the excess -- their money needs are uncertain, but pooling reduces the uncertainty of how much money the depositors in aggregate will actually withdraw at any point in time.

[Jussi]

Oliver,

I don't quite follow you. It seems to me that the discussion is now mixing theory and practice.

You were first asking: "How does money enter the non-bank economy without first being credit money?"

I took/take this as a theoretical question. In theory the answer is that the government / Central Bank can purchase something (from the primary or secondary markets), spend it into existence or just give out for free.

[Oliver]

Jussi

I specifically did not consolidate government and the central bank. I said the central bank can not buy in primary markets - the reason being that that is not its role, as opposed to what monetarists, and apparently you, claim. A theory that explains practice, as theories should :-).

[Too Much Fed]

Jussi's post said:

"@TMF:

The credit card itself is instructions to issue a "green" bond to the credit card issuer.

The red money itself is instructions to issue a "green" bond to the red money issuer?

The credit card is not "green" money, is not "red" money, and is not a "green" bond. The credit card is just instructions.

When you use a credit card at a grocery store, here is what happens. You swipe the card saying I want to actually borrow demand deposits ("green" money). You sell a new ("green") bond to the bank and buy new demand deposits from the bank. The bank sells new demand deposits to you and buys a new bond from you. These show up on both balance sheets.

Next, you sell demand deposits to the grocery store and buy food. The grocery store sells you the food and buys the demand deposits. These show up on both balance sheets.

Sound good?

[Too Much Fed]

David Andolfatto said:

"Nick,

I haven't had time to read that report or even the comments above, but I'll go ahead and ask anyway: Why would we want to permit overdrafts in the first place in a 100% reserve banking system? If people need to borrow money, there'd be a separate credit market on which they could do so. That's the idea behind separating money from credit.

David"

Exactly, an overdraft is actually a loan for an entity that would make a mistake on its check book so that the transaction is not declined.

I have $500 in demand deposits in my check book. I write a check for $1,000 in demand deposits by mistake. I have "overdraft coverage" (a loan in advance).

$500 of existing demand deposits is moved from my account. $500 of new demand deposits is placed in my checking account and moved from my account. I now have a new $500 bond liability that moves to the bank.

There is no "red" money. The "red" money is actually a "green" bond. The "green" bond has a high probability of NOT being used as a medium of exchange. It is not medium of account either.

[Jussi]

TMF,

"borrow demand deposits ("green" money)"

So my money account has a negative balance?

What happens is that I get a record saying I got short in demand deposits (red notes) and the store got demand deposit (green notes). Both were created by the bank almost in real time. This is how Nick once wrote red and green notes work:

"each goes to the central bank and asks for 5 green and 5 red notes, the buyer gives 5 green notes to the seller, the seller gives 5 red notes to the buyer, and they do the deal."

(http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/02/colateral-and-the-money-supply.html)

But you are quite right that technically I cannot sell anything to pass on my used credit forward - in that sense it is only "one time" medium of exchange and thus not true red money.

[Too Much Fed]

Jussi, http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/02/colateral-and-the-money-supply.html

"Now suppose we add a second form of central bank currency. Red currency has negative value. If you buy $10 worth of apples, either you give the seller 10 green notes, or the seller gives you 10 red notes, or some combination of the two."

I don't believe I have ever seen that supposition in the real world. There are green currency, green central bank reserves, green demand deposits, and green bonds.

"So my money account has a negative balance?

What happens is that I get a record saying I got short in demand deposits (red notes) and the store got demand deposit (green notes). Both were created by the bank almost in real time. This is how Nick once wrote red and green notes work:"

Let's assume your checking account (money account) can never have a negative balance (realistic). You can spend the existing green deposits you have. If you go over it, you have to borrow/have overdraft protection. At the end of July, you receive a statement that says you have $500 in "green" demand deposits in your checking account and "green" bond/loan balance of $0 owed to the bank.

On August 30th, you spend $1,000 in "green" demand deposits with your credit card (it may need to be a debit card). You got "short" $500 in "***green***" demand deposits by issuing a "green" bond to the bank (there is nothing "red"). Your statement on August 31st will say you have $0 in "green" demand deposits in your checking account and your "green" bond/loan balance will say $500, which is owed to the bank. The store got $1,000 in "***green***" demand deposits added to its checking account.

The $500 in new "green" demand deposits were created by the bank. The $500 in "green" bonds/loans were created by you (not the bank). The bank paperwork formalizes your "green" bond. The "green" bond/loan was created by you. The bank can not force you to issue the "green" bond (liability).

I am pretty sure this is how the accounting works.

You have a new $500 "green" bond that is both an asset and a liability to you. The bank has $500 in new "green" demand deposits that are both an asset and a liability to the bank. The overdraft triggers an asset swap. You sell the new $500 "green" bond and buy the $500 in new "green" demand deposits. The bank sells $500 in new "green" demand deposits and buys the new $500 bond/loan from you.

The demand deposits are always "green" whether borrowed or not. I hope I got everything in the right place.

[Oliver]

TMF
That is all so obvious it is hard to believe anybody could think otherwise.

[Jussi]

TMF,

"You got "short" $500 in "***green***" demand deposits."

http://en.wikipedia.org/wiki/Short_%28finance%29

"Mathematically, the return from a short position is equivalent to that of owning (being "long") a negative amount of the instrument."

[Too Much Fed]

Jussi, I am going to say "red" currency or "red" notes are actually "green" bonds/loans. Assuming that, the central bank can't force someone to issue those liabilities. It can't force entities into debt. The entities have to agree to issue "green" bonds/loans.

Nick gives the impression with his supposition that "green" bonds/loans ("red" currency) can be issued by the central bank at will. I do not think that is true.

Jussi said from above: "each goes to the central bank and asks for 5 green and 5 red notes, the buyer gives 5 green notes to the seller, the seller gives 5 red notes to the buyer, and they do the deal."

Let's change that around with "green" bonds/loans with the central bank acting like a commercial bank for loans. Each goes to the central bank and asks for 5 green notes and 5 bonds/loans, the buyer gives 5 green notes to the seller, the seller gives 5 bonds/loans to the buyer, and they do the deal.

They both went into debt to the central bank. The 5 green notes that go to the seller is a usual "money" transaction. The buyer takes over the 5 bonds/loans from the seller, an unusual transaction. At the end, the buyer has 10 bonds/loans owed to the central bank. The seller has 10 "green" notes. It is just as if the buyer went to the central bank and got a loan (10 loans/bonds owed to the central bank and 10 "green" notes) and then exchanged the 10 green notes for goods.

I can't see any good reason for the "red" currency or "red" notes. When looked at this realistic way ("green' bonds/loans), it does not appear that net money equals zero and the bonds/loans will probably not be used as medium of exchange, just like the real world.

The best thing to do is say no overdrafts. Then the loan/bonds part is easier to see.

[Too Much Fed]

Oliver said:

"TMF
That is all so obvious it is hard to believe anybody could think otherwise."

It must not be obvious or we would not be discussing this. I agree it should be obvious. It seems to me some other people do think otherwise. Explaining debt and banking should be obvious. It is not. Believe me.

Oliver, I think demand deposits and currency are both medium of account and medium of exchange. Do you agree or not?

[Too Much Fed]

Here is JKH talking about the ECB and the fed:

http://bilbo.economicoutlook.net/blog/?p=16898

"“Asset based” and “overdraft” are not very apt descriptors.

The significant difference between Fed (ab) and ECB (o) systems, is that the assets of the Fed are normally claims on government, and the assets of the ECB are normally claims on banks.

The exact nature of the claim is less important to the distinction. Most Fed assets normally are outright holdings of Treasury bonds. Most ECB assets normally are collateralized lending to banks."

"collateralized lending to banks" and "claims on banks" mean the banks are issuing a bond or bond-like instrument to the central bank (the ECB here). It is a "green" bond/loan for "green" currency and/or "green" central bank reserves.

[Roger Sparks]

TMF said
"The significant difference between Fed (ab) and ECB (o) systems, is that the assets of the Fed are normally claims on government, and the assets of the ECB are normally claims on banks."

I think this is correct in ordinary times.

Off topic, but am I correct that with the last QE, the Fed traded money (green) for Mortgage Backed Securities? Of course, the Fed has no money except what it prints or keys.

In the case of the ECB, am I correct that their current QE is trading euros for national bonds? Of course, the ECB has no euros except for what it prints or keys.

In Nick's terminology, I think both Central Banks are printing green notes and trading them for bonds. I don't see the dual green-red note trade here.

[Jussi]

TMF,

"I am going to say "red" currency or "red" notes are actually "green" bonds/loans."

I do not think that is close enough. It is critical in the context that red notes are medium of exchange.

I think the red notes were introduced ONLY as an analytical vehicle to gain insight how the models / proposals work.

[Oliver]

Oliver, I think demand deposits and currency are both medium of account and medium of exchange. Do you agree or not?

I would put it this way: in a two tier system with banks and a central bank, the cb's money is the medium of account. It is the benchmark with which demand deposits, give or take miniscule differences in interest paid / fees charged, trade at par. If you think of bank notes / demand deposits as bank bonds, then in a system without a unifying central bank that intervenes in the bank bond market, you would not expect an outcome where one bank's bonds automatically trade at par with another's. And I would say means of payment not medium if exchange, that has such a bartery ring to it.

[Too Much Fed]

Jussi said: "I do not think that is close enough. It is critical in the context that red notes are medium of exchange.

I think the red notes were introduced ONLY as an analytical vehicle to gain insight how the models / proposals work."

It seems to me the "red" currency or "red" notes give the impression they have no default risk, while they actually do have default risk just like "green" bonds. For example, with the 5-5 exchange. I go into debt to the central bank (5 "green" notes and 5 "red" notes). I give 5 "red" notes to the buyer of goods, give 10 goods to the buyer, and receive 5 "green" notes from the buyer. The buyer defaults on all 10 "red" currency/"green" bonds. 5 of those are my liability. I'm not going to do that scenario. The "red" currency/notes and "green" bonds won't be used as a medium of exchange.

I am assuming "red" currency/note = overdraft and overdraft = loan.

To gain insight into how the system works, it has to be described properly. I'd say calling the ECB an "overdraft" system with "red" currency/notes is not a very good way to describe it. I'd say calling the ECB system an "overdraft" system with "green" bonds/loans from the banking system to the ECB is better.

[Too Much Fed]

Oliver said: "I would put it this way: in a two tier system with banks and a central bank, the cb's money is the medium of account. It is the benchmark with which demand deposits, give or take miniscule differences in interest paid / fees charged, trade at par."

Do the demand deposits of each of the commercial banks along with each other and currency have a fixed conversion rate (that just happens to be 1 to 1) both ways?

[Too Much Fed]

Roger Sparks said: "Off topic, but am I correct that with the last QE, the Fed traded money (green) for Mortgage Backed Securities? Of course, the Fed has no money except what it prints or keys."

I don't remember which QE it was.

What if the fed earns $50 billion in interest (existing "green" money) from its assets and then uses that to buy MBS?

[Roger Sparks]

TMF said

"What if the fed earns $50 billion in interest (existing "green" money) from its assets and then uses that to buy MBS?"

I think the scale of purchases is vastly larger than the interest the Fed earns. From the Federal Reserve Data Series 'MBST', the purchases were about $1.7 trillion over 5 years.

[Oliver]

Do the demand deposits of each of the commercial banks along with each other and currency have a fixed conversion rate (that just happens to be 1 to 1) both ways?

What's your question? Is the exchange rate 1:1 by chance? No, it's a peg to the medium of account. One could also say a franchise of the medium of account. The important thing being that the cb does not control the amount of means of payment in the (non-bank) economy. The cb just manages the franchise by homogenising a tranche of its member bank books, thus enabling a smooth payment system without runs and interest rate spikes. All a very complicated way of explaining endogenous money, I guess.

[Oliver]

TMF
from (via rwer): http://eng.forsaetisraduneyti.is/media/Skyrslur/monetary-reform.pdf

The CBI has monopoly on issuing notes and coin. Coin is manufactured for the CBI by the Royal Mint, and notes by a specialist printer in the United Kingdom.
Banks may purchase new coin or notes from the CBI in return for central bank reserves or securities. Individuals and firms cannot buy notes and coin directly from the CBI, only from banks, in exchange for a reduction in the balance of their deposit account.
...
An important function of the CBI is to be the ‘banker to the banks’. This involves providing commercial banks with accounts for holding central bank reserves. These reserve accounts allow commercial banks to make payments to each other by transferring reserves between their respective accounts at the CBI.
In addition to issuing money and providing reserve accounts for banks, the CBI provides a number of bank accounts to the government, in which funds from taxation and borrowing are temporarily held, before being used for government spending or paying the interest on previous borrowing. Among the CBI's other duties is the setting of monetary policy (through the policy rate of interest), promoting price stability, promoting financial stability, maintaining foreign exchange reserves, and operating a domestic payment system and payments abroad.
...
In addition to issuing money and providing reserve accounts for banks, the CBI provides a number of bank accounts to the government, in which funds from taxation and borrowing are temporarily held, before being used for government spending or paying the interest on previous borrowing. Among the CBI's other duties is the setting of monetary policy (through the policy rate of interest), promoting price stability, promoting financial stability, maintaining foreign exchange reserves, and operating a domestic payment system and payments abroad.

[Frank Restly]

TMF,

"I can't see any good reason for the red currency or red notes."

One reason for the red money is to suspend property rights. Think of it this way:

With a single green money system - property rights are enforced in that the holder of green money has a legal protection to not be robbed of his / her positively valued assets. With both green and red money, property rights can be fully extended for green money and be nonexistent for red money. If someone steals your red money, do you really care?

Why would someone steal your red money? Because he / she wants to sell a good and can't find enough buyers with green money, so he / she steals your red money and then "sells" the good by transferring both the good and the red money to the buyer.

Sort of "Through the Looking Glass", but the implications are interesting.

[Oliver]

Frank

With a single green money system - property rights are enforced in that the holder of green money has a legal protection to not be robbed of his / her positively valued assets. With both green and red money, property rights can be fully extended for green money and be nonexistent for red money. If someone steals your red money, do you really care?

The point about green money is that the value of the green asset is identical to the obligation inherent in the negative green debt. My asset is only worth something because somebody owas it to me. If debt (negative green money) is an IOU, then money (green money) is an UOI. It has value because I believe that U will honour your debt to I. The whole green and red discussion is based on the notion that U and I can be separated. It postulates UOs and IHaves. Problem is, I can only Have (and thus legally enforce) what UO.

[Nick Rowe]

Csissiko: "These proposals are premised on the idea that banks are "just" intermediaries between savers and lenders, so nothing will be lost by forcing debt out of the payments system."

I think it's more the idea that we can divide banks into two parts: one part is just an intermediary between borrowers and lenders; the second part handles the payments system.

[Frank Restly]

Oliver,

"The whole green and red discussion is based on the notion that U and I can be separated."

It goes further than that because of this note under Nick's #3 - "Because red notes are media of exchange too."

We need to accept that red notes are not only an obligation, but also a "negative money" in the sense that they circulate through an economy in the opposite direction that green money does. Green money goes from buyer of good to seller of good, red money goes from seller of good to buyer of good. Red money provides an additional method of allowing markets to clear. Legitimized thievery of red notes enhances this process.

We could drop the red note / green note connotation, and think of it in terms of money and debt, but this would not be your parent's debt. This would be debt that is significantly more liquid than what exists today and would be uniform in construction - something like perpetual debt that is passed down to future generations. There would be no 30 year debt versus 2 year debt. Debt of the modern variety is often extinguished when the borrower dies (personal debt) or goes bankrupt (personal or corporate debt), while the money borrowed money can remain. Debt of the "red money" variety would presumably be passed along to inheritors or spread among the remaining citizenry after a person / company dies or goes bankrupt.

[csissoko]

Nick: "I think it's more the idea that we can divide banks into two parts: one part is just an intermediary between borrowers and lenders; the second part handles the payments system."

Fair enough. I should revise that statement to read: "These proposals are premised on the idea that when banks lend they are "just" intermediaries between savers and lenders, so nothing will be lost by forcing debt out of the payments system."

The point that I was trying to make is that these proposals assume away the possibility that an important economic role is played by banks that source the funds they lend by expanding the money supply.

[Jussi]

"nothing will be lost by forcing debt out of the payments system."

I do not have the answer but I think there is something in this and it should be discussed thoroughly.

Robinson and Friday were living on an island. They can just barter - a classical economy. Robinson wasn't too keen on making any changes and he decided he will not invest at all (this was a econparadise so no depreciation either). Robinson owned everything and Friday couldn't make any invest by himself as Robinson paid so little (just an assumption) to him that he needed to consume it all.

Then Ben came around and started a bank with his tiny capital (Ben was first giving away his notes for free to get his money adapted!). Friday talked to Ben who also liked the investment possibilities. Together they decided to go ahead and make the investments far greater than Ben's bank capital. Ben funded it with his fountain pen - out of thin air. But this way the resources for the investments were mostly forced out of Robinson's consumption - without his consent and through inflation if necessary. And the cost of bad investments will also be beared by Robinson; only profits are not coming his way. And all this without explicitly violating Robinson's ownership rights!

Robinson foresaw all this and came up with a stagnant money supply proposal to avoid systemic instability that was threatening their paradise. After it Ben would need Robinson's money to be willingly put on investment account before he could fund any investments.

I'm not sure what is the moral of the story.

[csissoko]

Jussi: "the investments were mostly forced out of Robinson's consumption - without his consent"

This is where you lose me. Robinson will not accept Ben's banknotes in payment unless there's something in it for him. If there is, then the banknotes are promoting socially valuable investment (on the Pareto principle). If there's nothing in it for Robinson, the banknotes will never circulate.

[csissoko]

For me the moral of the story is that at least in a three-person economy, we can expect the innovation of banking to improve, not reduce the set of possible outcomes. (Throw in asymmetric information and a few hundred more people and I would be much more cautious about this claim.)

[Frank Restly]

Nick,

My thoughts on taxation in a red / green currency economy:

Government is permitted to spend only the net of green notes that it receives. It automatically returns equal amounts of red / green notes that it receives in taxes to the central bank to be destroyed / reintroduced into the monetary system. When a government receives in taxes more red notes than green notes, it either redistributes those red notes across the population or returns the excess to red money tax payers.

I would think that treating red and green notes as equals from a tax policy view has an advantage over taxation of green currency only. It provides a means for the central bank to reduce the overall supply of liquidity (matching red and green notes) without suspending green note property rights and without introducing a fixed lifetime for red note / green note pairs.

A traditional debt / money system has a timeframe over which both the money and debt exist and then are retired (private debt only, government debt with rollover complicates this). A red money / green money system would need a taxing government to give the central bank the opportunity to remove total liquidity (equal parts red and green notes) at any point without suspending green money property rights.

And so from an balance sheet sense, green notes represent an asset of the owner where red notes represent a liability of the owner. But from a liquidity and taxation sense, red notes and green notes both represent an asset. Both red notes and a green notes could satisfy payment on taxes and both red notes and green notes could be used to facilitate a trade of goods.

[Jussi]

"This is where you lose me. Robinson will not accept Ben's banknotes in payment unless there's something in it for him. If there is, then the banknotes are promoting socially valuable investment (on the Pareto principle). If there's nothing in it for Robinson, the banknotes will never circulate."

Yes, I guess you are correct but only in this three men setup.

If we introduce an ownership class of Robinsons and worker class of Fridays, which both consist of enough people to make barter inefficient enough, then Robinsons shouldn't refuse Ben's notes on Pareto basis and they still be forced to pay the investments they didn't want in the first place. And I think there isn't necessary a (Pareto) limit for that because individually they cannot fall back on barter even if they would be collectively better off (network effect).

I think you are right that toy examples might not help us to understand how the banking works. For me it is interesting that in a simple setup banking might give us more investments and less inequality.

[Nick Rowe]

csissoko: "The point that I was trying to make is that these proposals assume away the possibility that an important economic role is played by banks that source the funds they lend by expanding the money supply."

I think I tend to agree. Here is a recent post which looks very different from your two most recent posts, but which I think is heading towards the same general conclusion.

Another way of saying the same thing (I think): it is not (generally) optimal to have a world with only green money. There are two quantities that matter:

1. Green plus red money
2. Green minus red money

[csissoko]

Nick, this is how I understand your post on fractional reserve banking, OQM, and 100% reserve banking:

You (and Friedman) are doing your analysis in a complete markets world, but the concept of liquidity is an incomplete markets concept. Because of distributional effects, it is far from clear that the Friedman Rule is generally optimal (or Pareto improving) in an incomplete markets world (where money has value for transactions purposes).

This is how I understand the situation (and I may be guilty of failing entirely to understand Friedman). In a complete markets model, liquidity is already perfect. Thus, to the degree that some contrivance is introduced into the model to induce people to hold a low return asset such as money, the ideal policy is to drive the return of that asset up so that it equals the return on the many other sources of liquidity in the model.

I guess I would say that the whole discussion is about an aversion to dealing with the incomplete markets models that are necessary to explain why people hold money. The way I understand your results in your post is that you are coming face to face with the contradiction that is the (useless) introduction of money into a complete markets model.

I am not entirely clear on how you got from the post to "Another way of saying the same thing (I think): it is not (generally) optimal to have a world with only green money.", but I agree entirely with the latter statement. "green money plus red money" is what makes it possible for banks to lend up to the needs of the real economy, and this is much more important than the price of green money.

[Nick Rowe]

csissko: "You (and Friedman) are doing your analysis in a complete markets world, but the concept of liquidity is an incomplete markets concept."

I only half agree with you there. The Old Monetarists would agree with you that the reason why people use money, and the reason why people hold money, depend on some fundamentals, and you might say those fundamentals are "incomplete markets" (or, at least, some costs of transacting in a way that would make markets complete). Now the Old Monetarists did not build those fundamentals explicitly into their models, in the way that New Monetarists attempt to do (and are to be applauded for doing). It is only because we use money that the assumption of as if complete frictionless markets does not fall flat on its face; but if markets really were complete and frictionless we wouldn't need to use money. The Old Monetarists waved their hands, but they weren't daft.

The question is: if we do model those fundamentals explicitly, does that really overturn the Friedman Rule in any important way? Do we need to model why people hold stocks of refrigerators to say that a monopolist producer of refrigerators should price them at marginal cost for optimality?

"I am not entirely clear on how you got from the post to...."

Nor am I entirely clear :-) I'm still thinking this through.

[Duncan Cameron]

Nick would you be good enough to react to this new Bank of England study? It certainly takes a big step towards the money is an IOU perspective. Thanks. http://www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q102.pdf

[csissoko]

Nick,

I certainly don't think the Old Monetarists were daft. I think they developed their intuition from flawed models, and then were over-committed to those models. When it comes to the Friedman Rule, the more I think about it the more that I conclude that the problems with it are less related to complete markets and derive more from a failure to explore the implications of heterogeneous agents carefully.

"if we do model those fundamentals explicitly, does that really overturn the Friedman Rule in any important way?"

First, it's not even clear that we need to model fundamentals explicitly to overturn the Friedman Rule. Heterogeneity of agents alone may be sufficient: https://www.econ.iastate.edu/research/journal-articles/p1852

Second, your intuition is completely correct that, given a representative agent or its equivalent, even when fundamentals are explicitly modeled, the optimum return on a green money-type instrument will follow the Friedman Rule. (e.g. Wright and Williamson find that Friedman Rule holds here: http://www.artsci.wustl.edu/~swilliam/papers/methodsstl.pdf )

Thus, the real issue is that given heterogeneous agents it will often only be possible to implement the Friedman Rule by treating every different type of agent differently. (See here for a fairly general treatment: https://www.econ.iastate.edu/research/journal-articles/p1827 ) So the conclusion I draw from the analysis of the Friedman Rule is that optimal government management of "green money" in the real world with heterogeneous agents is not feasible -- and that we need to think about other types of money.

To me the issue is less one of overturning the Friedman Rule than of reconceiving what we mean by the term "money", and reducing the focus on "green money" which is so closely tied to the Friedman Rule. In this sense, I think we're very much on the same page.

"Do we need to model why people hold stocks of refrigerators to say that a monopolist producer of refrigerators should price them at marginal cost for optimality?"

Money is very different from other goods, because liquidity constraints so often define people's opportunity sets -- and therefore the difference between realized GDP and potential GDP. While it is clear that a "good" (or green money) can be used to address liquidity constraints, it is far from clear that goods are the only way to address liquidity constraints, and thus it far from clear that we should be analogizing to goods at all when we talk about money.

[Nick Rowe]

Duncan: I did a blog post on it soon after it came out.

A financial asset is just a bit of paper (or plastic, or silicon nowadays) with some sort of IOU (or promise) written on it. Nowadays (nearly?) all money is IOUs (but not all IOUs are used as money). Bank of Canada currency is an IOU, but it's a strange sort of IOU. The promise (though it's written on the BoE's website, not on the currency itself) says (roughly): "We promise to make this bit of paper/plastic depreciate at roughly 2% per year against the CPI basket, unless we change our minds. Signed Steve and Carolyn."

csissko: imagine a world where hunters get lucky or unlucky in catching deer. Lucky hunters have a big demand for fridges, and unlucky hunters have a small demand for fridges. And the government is the monopoly producer of fridges. It might be ex ante Pareto Optimal for the government to price fridges above marginal cost, and redistribute the monopoly profits as a lump-sum transfer. It creates insurance. I *think* that might be what's going on in those heterogenous agent models. But I don't see that as a *fundamental* critique of marginal cost pricing. It's a bit like the old argument that the optimal inflation tax is positive, because the government can't tax what it can't observe (were the hunters who caught no deer unlucky or lazy?) and tax evaders use currency.

[Frank Restly]

Csissoko,

"While it is clear that a good (or green money) can be used to address liquidity constraints, it is far from clear that goods are the only way to address liquidity constraints, and thus it far from clear that we should be analogizing to goods at all when we talk about money."

A good as a means to satisfy liquidity constraints can be durable, it can be portable, it can be easily counted / divisible, it can be uniform in construction, and the boundaries for the protection of property rights can be well defined.

That doesn't mean that a nongood means of addressing liquidity constraints cannot be conceived of, but I would think there would be trade offs to consider.

I suppose that information could be used to address liquidity constraints.

Information is durable (with a recording mechanism), it is portable (with a recording mechanism or by word of mouth), and it is countable / divisible (units are bits / bytes).

Information is also non-uniform and property rights for information can be difficult to define.

[csissoko]

Nick:

re: heterogeneous agent models. I don't think I was clear. These models do not constitute a fundamental critique of marginal cost pricing -- that result still stands in the models at which I have looked closely.

The critique is in the informational demands of implementing marginal cost pricing of "green money." The problem with monetary policy is that it's not just about the price, but also about the taxation regime that supports the monetary policy. In particular, low-information lump-sum taxation (equal across all agents) cannot be used to implement marginal cost pricing. To implement marginal cost pricing, the monetary authority needs high-information taxation that discriminates across all the different types of agents.

In a heterogeneous agent model, assume a Friedman Rule rate of interest on current money holdings and lump sum taxation. If the tax is such that the high (liquidity) needs types has enough cash, then it is necessarily the case that the low needs individual has too much cash -- and will choose to purchase more than optimal bundle of goods. So we don't have an equilibrium. If the tax is such that the high needs type has too little cash, then that type can't purchase the optimal bundle of goods. The problem is that fiat money is not a good way of allocating liquidity, so any lump-sum taxation equilibrium has to be a second-best equilibrium. The alternative is an all-knowing monetary authority that can redistribute fiat money to implement the allocation associated with the Friedman Rule.

In short, in heterogeneous agent models marginal cost pricing is still optimal for fiat money, it's just not obvious how a monetary authority can put marginal cost pricing of fiat money into place.

I read this theoretic problem to mean: we need banking, aka debt-based money, or in your terminology green and red money.

Frank: "That doesn't mean that a nongood means of addressing liquidity constraints cannot be conceived of, but I would think there would be trade offs to consider. I suppose that information could be used to address liquidity constraints."

We're in luck, because our forebears spent the past few centuries developing banking, which as far as I am concerned is mechanism for converting information into money. So we have the solution. All we need to do is understand how it works, so we don't destroy the mechanism.

Yes there are trade-offs to information-based money: managing inflation (much easier under a gold standard), and avoiding financial instability (mass bank failures). On the other hand, we've had a reasonably well-functioning bank-based monetary system for decades (if not centuries), so we can probably manage these problems if we work hard at understanding them.

[Nick Rowe]

csissoko: "If the tax is such that the high (liquidity) needs types has enough cash, then it is necessarily the case that the low needs individual has too much cash -- and will choose to purchase more than optimal bundle of goods."

It's the second bit of that sentence I don't get. If cash pays sufficiently high interest, why would the individual who is satiated in cash (relative to other assets) choose to purchase more than the optimal bundle of goods? For example, if the government produces two types of asset: bonds and cash, why shouldn't the government set the rate of interest on cash equal to the rate of interest on bonds, so that people are indifferent between the two? Or, more simply, why produce any bonds at all? Produce cash only, and set the interest rate on cash high enough so that aggregate demand is just right?

[csissoko]

Nick: I think what's going on is that with heterogeneous agents, is that the economy's aggregate transversality condition is not enough. It is equally the case that each type of agent cannot reasonably have a path of money holdings that is forever growing, but never spent (which is how I understand what you are proposing). Any agent who looks into the infinite future and expects in every future period to hold more than enough fiat money to buy the "optimal" consumption bundle will if he is self-interested decide to purchase more goods in every period instead of buying the optimal bundle -- since you can't eat fiat money.

The underlying problem is that the price that is "just right" for aggregate demand, is not "just right" for each type of individual in the economy -- unless the monetary authority uses tax policy to adjust everyone's money holdings in every period.

[Frank Restly]

Csissoko,

"Any agent who looks into the infinite future and expects in every future period to hold more than enough fiat money to buy the optimal consumption bundle will if he is self-interested decide to purchase more goods in every period instead of buying the optimal bundle -- since you can't eat fiat money."

And if the same agent looks into the infinite future and expects that all his neighbors in every future period will be able to hold more than enough fiat money to buy the optimal consumption bundle, will he refuse to accept his neighbor's fiat money for goods? If I am satiated in liquidity (and always will be) and my neighbor is satiated with liquidity (and always will be) why would either of us trade goods for money?

Perhaps because I am selling apples that take six months to produce and he is selling airplanes that take two years to produce? There are still production schedules to consider. Does "liquidity for all" protect me if I deliver my apples and then wait another 18 months for his airplane? How does overly abundant liquidity mitigate the exchange of two goods (apples and airplanes) that have different production time frames.

I tell my neighbor I will sell him 1 million of my apples for his 1 airplane. Six months later, I give him 1 million apples, he gives me 1 million green notes. Eighteen months later I give my neighbor 1 million green notes, he gives me the 1 million green notes back and I don't receive an airplane. I go to government asking for my airplane, government tells me "Here is some interest on those 1 million green notes".

[csissoko]

Nick: Correction to my previous (1:57) comment. The fact that the transversality condition is violated for the low needs type implies that the aggregate transversality condition is violated. So the basic problem is that a government lump-sum tax policy that meets the needs of the highest type results in a violation of the transversality condition (as the low needs type carries ever increasing amounts of money).

Frank: Precisely. The discussion is about why satiation is not an equilibrium phenomenon.