Media of exchange and the clearing house

What makes our chequable demand deposits at banks media of exchange? The answer is: the clearing house, where circles of offsetting IOUs are cancelled out.

It is clearing houses that create money. Unless there is only one bank, where everybody banks, because it doesn't need a clearing house.

Nothing revolutionary here (except maybe a little bit right at the end). This is just me trying to get my head straight on something, following a discussion with JKH in the comments.

1. Amanda cuts Betty's hair for $20. Betty manicures Cathy's nails for $20. Cathy massages Amanda's back for $20. They pay for each other's services with currency. That currency is their medium of exchange. It flows around the Wicksellian triangle in the opposite direction to the flow of services. $60 in total transactions with currency.

2. Suppose Betty is temporarily short of currency, and gives Amanda an IOU for $20 in exchange for the haircut. And then pays off (buys back) her IOU a couple of days later for $20 in currency (maybe plus interest). I say that Betty's IOU is not a medium of exchange. Betty has merely postponed payment. Betty still needed to pay Amanda $20 in currency. Being able to buy on credit may reduce Betty's demand for medium of exchange, but it is not itself a medium of exchange. $60 in total transactions with currency.

3. Suppose Betty is temporarily short of currency, and gives Amanda an IOU for $20 in exchange for the haircut. Suppose Amanda gives Betty's IOU to Cathy in exchange for a manicure. And suppose Cathy gives Betty's IOU to Betty in exchange for a manicure. Betty then tears up her IOU to herself. They didn't need any currency at all. Betty's IOU was used as a medium of exchange. Betty is like a central bank, because everyone accepts her IOUs. $0 transactions in currency.

4. Suppose all three women are temporarily short of currency, and all three give their IOUs in exchange for services. And all three repay those IOUs with currency a couple of days later. There were still three transactions of $20 for currency. Again, I say that those IOUs are not media of exchange. $60 in total transactions with currency.

5. Suppose all three women are temporarily short of currency, and all three give their IOUs in exchange for services. After a couple of days, all three women meet together, in a place called a "clearing house". They add up each woman's debits and credits. Each has a credit of $20 and a debit of $20. They agree to cancel out the circle of offsetting debits and credits. They tear up the IOUs. They didn't need any currency at all. I say their IOUs are now media of exchange. $0 transactions in currency.

In that 5th case, what makes their IOUs media of exchange is the clearing house.

Now suppose that each of the three women starts a financial intermediary. She both borrows and lends. She sells her own IOU for $20 (she borrows $20), and uses that $20 to buy someone else's IOU (she lends $20).

Are their IOUs media of exchange? Are their financial intermediaries banks, that create money? That depends.

6. (same as 3). Suppose Bill uses Betty's IOU to buy carrots from Chris, who uses it to buy apples from Andy, who uses it to buy bananas from Bill. Then Betty's IOU is a medium of exchange. Nobody needed to use any currency at all. Betty runs a bank. All three men bank with Betty. $0 transactions in currency.

7. Suppose Bill uses Betty's IOU to buy carrots from Chris. Chris uses Cathy's IOU to buy apples from Adam. And Adam uses Amanda's IOU to buy bananas from Bill. But the boys don't like holding IOUs from girls with the wrong first letter in their names. Each boy banks with his own girl. Adam goes to Amanda and swaps Cathy's IOU for another one of Amanda's IOUs. Bill goes to Betty and swaps Amanda's IOU for another one of Betty's IOUs. Chris goes to Cathy and swaps Betty's IOU for another one of Cathy's IOUs. What happens next?

7a. (same as 2.) If the three women don't have a clearing house, Amanda gives Betty $20 to repay her IOU, Betty gives Cathy $20 to repay her IOU, and Cathy gives Amanda $20 to repay her IOU. You get as many currency transactions as if the three guys had used currency in the first place. The women's IOUs are not media of exchange. $60 in total transactions with currency.

7b. (same as 5.) The three women meet after a couple of days and cancel their IOUs in the clearing house. Nobody needed any currency at all. Their IOUs are media of exchange. The three women's financial intermediaries are banks. They create money. $0 transactions in currency.

But notice something that I fudged. The three women meet every "couple of days" and cancel out circles of offsetting IOUs in the clearing house. It matters a lot how frequently the clearing house meets.

Assume a stationary economy, but where the timing of discrete transactions is random. In one limit, as the clearing house meets less and less frequently, the percentage of gross transactions that need to be settled with currency approaches 0%, because they all net out. (It wouldn't be stationary if it didn't). But in the other limit, as the clearing house meets continuously, the percentage of gross transactions that need to be settled with currency approaches 100%, because none of them net out.

The extent to which chequable demand deposits are media of exchange depends on the percentage of transactions that net out in the clearing house, which is a negative function of how frequently the clearing house meets. In the limit, with continuous clearing, only one transaction is settled at a time, and none net out. All have to be settled with currency (i.e. reserves). So chequable demand deposits are not media of exchange, in the limit. But the velocity of circulation of reserves increases towards some very large number, as you approach that limit.

I think that's right.

Comments

[Determinant]

Wow, a Nick post I agree with 100%.

[Nick Rowe]

Determinant: well, there's no religion or politics in it! And the only math is basic addition. (Though, there is a bit of implicit stats at the end.) I'm wondering if it has any policy-relevance at all.

[Odie]

Do you mean with "stationary economy" one that meets the definition of the Wicksellian Triangle where all goods and service transactions net out to zero? Would that not quickly be violated when just one girl such as Cathy decides to do her manicure herself and just save the other girls IOUs? Then the IOUs may never cancel out because Cathy keeps a "surplus".

"The extent to which chequable demand deposits are media of exchange depends on the percentage of transactions that net out in the clearing house, which is a negative function of how frequently the clearing house meets."

Not sure if I got the last paragraph correctly but when I write a check to someone who happens to have his account at the same bank I don't see a reserve transaction but would still believe my deposit functions as MOE.

[Peter N]

Clearinghouses can also supervise their members and act as lenders of last resort. They did this in the days before the Federal Reserve, whose formation was motivated by the failure of this system to stop the panic of 1907 (It took J.P. Morgan's personal intervention, diplomacy and occasional bullying of all concerned - a fascinating story).

From Wikipedia:

"Between 1853 and 1913, the nation experienced rapid economic expansion as well as ten financial panics. One of the Clearing House’s first challenges was the panic of 1857. When the panic began, leaders of the member banks met and devised a plan that would shorten the duration of the panic–and more importantly, maintain public confidence in the banking system. When specie payments were suspended, the Clearing House issued loan certificates that could be used to settle accounts. Known as Clearing House Loan Certificates, they were, in effect, quasi-currency, backed not by gold but by discounted county and state bank notes held by member banks. Bearing the words “Payable Through the Clearing House,” a Clearing House Loan Certificate was the joint liability of all the member banks, and thus, in lieu of specie, a most secure form of payment.
The certificates appeared in smaller denominations during the panic of 1873, and continued to be used as a substitute currency among the member banks for settlement purposes during panics in subsequent decades, including the Panic of 1893. Although they represented a potential violation of federal law against privately issued currencies, these certificates, as a contemporary observer noted, “performed so valuable a service…in moving the crops and keeping business machinery in motion, that the government…wisely forbore to prosecute.”

[Nick Rowe]

Odie: a stationary economy would be one where all banks are staying the same size, so one isn't gaining depositors from others, for example.

If two people have demand deposits at the same bank, that's case 6.

Peter N: sounds like that clearing house was acting very much like a central bank.

[Too Much Fed]

Nick said: "If two people have demand deposits at the same bank, that's case 6."

Are you saying Betty's IOU's and demand deposits function in the same manner?

[Peter N]

" sounds like that clearing house was acting very much like a central bank"

Though the Fed does some things the clearinghouse didn't and has somewhat different concerns. The Federal Reserve system was an evolutionary step. I doubt anyone today would model a central bank after it. OTOH the ECB model is equally strange and works a good deal worse.

I don't know whether you saw the post in Alphaville showing what a combined Fed-Treasury account would look like. It's very informative. Certainly students would benefit from seeing it.

It clearly makes the point that the Fed's reason for existence isn't one of financial structure, but rather a political necessary. The balance sheet results for monetizing the deficit and the government just creating and spending the money are identical. The politics seem to be quite different.

Political necessity is just as necessary as any other kind.

[JKH]

So if BMO and TD match $ 1 billion in cheques against each other in a single clearing ($ 2 billion gross total), but TD clears an additional net $ 50 million against BMO, you would say that the medium of exchange is cheques for $ 2 billion and central bank reserve balances for $ 50 million?

If so, I’m clear on the mechanics, but I’m not clear on the intuition.

I would think of medium of exchange here at two different levels – cheques for bank customers and reserves for banks.

And I think I would say for all of the transactions represented in this example that there is $ 2.050 billion in customer medium of exchange and $ 50 million in bank medium of exchange.

That seems to cover all of the payments that occur at both levels.

Maybe I’m misinterpreting what you’re saying though.

It sounds like you’re differentiating different media of exchange as that applies at the level of banks, whereas I’m thinking of different media of exchange at two different levels – banks and their customers.

In my example, I can see $ 50 million of reserves being a medium of exchange at the bank level, but the same $ 50 million in underlying cheques being a medium of exchange at the bank customer level (along with the other $ 2 billion).

I guess I’m not clear on that part of your framework – i.e. are you differentiating between the two different levels of banks and their customers? It doesn’t seem so.

But I’m not sure about all this.

E.g. what happens if you apply the same netting perspective at the level of an individual bank account? Is the medium of exchange the net cheque flow rather than gross in that case?

It really is a matter of setting out a definition.

[ATR]

Why can't you consider there to be multiple media of exchange associated with one transaction, if both people and banks are involved? Suppose I want to buy an apple from you. You will accept that transaction if you see your deposit account increase (or if I give you cash, but that's not as interesting). Your bank will accept that transaction if they see their clearinghouse IOU account or reserve (cash) account increase. Deposits or cash are media of exchange for you and me. Clearinghouse IOUs or reserves (cash) are the medium of exchange for banks. But both media need to be exchanged for that transaction to occur.

Also, not sure if this changes any of the thinking substantially, but I think it's best to think of some of these transactions as swaps of IOUs, not just 1 way IOUs. For case 2, you can think of Amanda giving Betty a loan, which involves Amanda giving Betty an IOU as well (just like a bank loan that creates a deposit, which is an IOU to pay cash).

Amanda assets: 20 IOU from Betty; liabilities: 20 IOU to Betty
Betty assets: 20 IOU from Amanda; liabilities: 20 IOU to Amanda

Betty uses her "20 IOU from Amanda" to pay Amanda for the haircut today, extinguishing her asset and Amanda's liability. She later pays the loan back with either another loan or cash.

[ATR]

Ah - I see JKH already suggested this.

Perry Mehrling views these sorts of things through a "hierarchy of money." Depending on the transaction and who is transacting, different forms of money are acceptable as media of exchange. And what may be viewed as 'money' at one level is seen as 'credit' at another.

[Too Much Fed]

ATR said: "Deposits or cash are media of exchange for you and me."

Now we are getting somewhere. Since demand deposits and currency (cash) are 1 to 1 convertible, are demand deposits and currency (cash) also medium of account (MOA)?

"Clearinghouse IOUs or reserves (cash) are the medium of exchange for banks. But both media need to be exchanged for that transaction to occur."

Even better. If someone withdraws currency and then deposits it in another bank, will the accounting end up the same as just "transferring" the demand deposits and central bank reserves?

[Too Much Fed]

ATR, if I bank at the same bank as a retailer, can I swap a demand deposit(s) for goods (meaning the transaction is settled) without monetary base (currency or central bank reserves) being involved?

[ATR]

TMF: Yes. The bank doesn't need any medium of exchange for the transaction to settle. But the retailer definitely does.

[Too Much Fed]

Odie said: "but when I write a check to someone who happens to have his account at the same bank I don't see a reserve transaction but would still believe my deposit functions as MOE."

I'd say so.

ATR, now if someone could convince Scott Sumner of that.

ATR, do you think currency and demand deposits are BOTH MOE and MOA?

[J.V. Dubois]

Hmm, another post of MoA vs MoE at least in essence. But here I think that you may be on to something. I *think* that this Wickselian clearing house is behind "money in the utility function" model. Market transactions are just somehow cleared utilizing money in the process. The value of this transaction service is equal to the loss that whoever ends up holding money is willing to shoulder. This transaction service is just another one from thousands of services being purchased on the market.

These are the moments where I think that Scott's MoA view of money may help steer the discussion back where it belongs. Because I think that considering or measuring the impact of sticky prices/wages on the amount of trade being made may be better in estimating the loss from unrealized trade then just assuming that it is equal to loss from holding money. I don't know, I have to think about it more.

[Nick Rowe]

JKH: I think you are interpreting me correctly.

What I am trying to get at is the distinction between:

1. $60 worth of goods get bought and sold, and they are bought with IOUs (credit), but you still need $60 of transactions in currency to make it all happen.

2. $60 worth of goods get bought and sold, and they are bought with IOUs (credit), and you don't need any transactions in currency to make it all happen.

Plus all the intermediate cases in between where you need some transactions in currency to make it all happen, but less than $60.

I want to say that the IOUs are 0% media of exchange in case 1, 100% media of exchange in case 2, and some percentage in between in the intermediate cases.

I'm not sure whether this is a useful definition.

I used to think that reserves are the medium of exchange for banks. And that is true, but it leaves out the important role of the clearing house, which reduces (and in the limit eliminates) the need for any payments using reserves.

I think my case 5 is the most interesting. There are no banks, but there is a clearing house. And the clearing house eliminates (more generally, reduces) the need for any transactions using currency. If we add in banks, we get case 7b, but nothing important changes. It is the three women's ownership of a clearing house, rather than their ownership of banks, that makes their IOUs function as a medium of exchange.

[Nick Rowe]

Peter N: "I don't know whether you saw the post in Alphaville showing what a combined Fed-Treasury account would look like."

I haven't see it, but I don't think there would be anything new there. Macroeconomists very frequently write down a consolidated budget constraint for the government and central bank: G + rB = T + deltaM + deltaB. In fact, writing down the consolidated budget constraint is more common than writing down two separate budget constraints.

JV: "I *think* that this Wickselian clearing house is behind "money in the utility function" model."

I don't think so. You see, I have ignored the stock/flow distinction in this post. In case 1, for example, there are $60 worth of transactions with currency, but over how long a period, and how big an average stock of currency are the three women holding? The velocity of circulation is undefined in that example. Even if you assume that people use currency for all transactions, that tells you nothing about the average stocks of money held, and that they want to hold.

What "money in the utility function" is trying to capture is the convenience of not having to coordinate your expenditures and receipts of money exactly, by holding an inventory of money.

[JKH]

Nick,

Starting at bank reserves:

“I used to think that reserves are the medium of exchange for banks. And that is true, but it leaves out the important role of the clearing house, which reduces (and in the limit eliminates) the need for any payments using reserves.”

Example:

BMO receives 100 net payment from TD in the clearings.

BMO pays 50 net to Royal.

BMO flat with all the other banks.

Then BMO receives 50 net multilateral from all other banks.

That’s the change in the level of BMO’s reserves.

So BMO receives 50 reserves as medium of exchange in the clearings.

The clearinghouse sorts out all the bilateral clearings so that bilateral clearings are all accounted for in the multilateral result for each bank - so there are no bilateral exchanges for each bank as such – just bilateral calculations that become embedded in the total multilateral number to be cleared for each bank.

(This is simplified for cheques – it’s the way things used to be at least with a single new reserve account number at the start of each business day - there are interfacing real time settlement systems now, I believe, and I’m just not sure how the timing works there in terms of bilateral payments that may hit the reserve account during the day (or not). In any event, I think it’s each real time change in the reserve account due to clearings of one sort or another that corresponds to the use of reserves as a bank medium of exchange.)

[Nick Rowe]

JKH: Yes. The way I see it is like this: with real time (continuous time) clearing, there would only ever be one payment at a time hitting the clearing house. So every single cheque for $100 from a BMO customer to a TD customer would cause BMO reserves to fall by $100 and TD reserves to rise by $100. There is no netting out at all. But reserves are flowing back and forth between banks very very quickly (extremely high velocity of circulation of reserves). And at the end of the "day", most changes in reserves would have netted out, cumulatively.

[JKH]

Nick,

There’s an interesting application of this to the government’s account at the central bank, which operates more or less like a bank reserve account.

Netting in bank reserve accounts – and in the government account – is an indicator of cash management discipline. Netting is the inevitable result of actually managing a cash position with the objective of being reasonable close to “flat” on a daily basis. The government does that too with its account. It tries to match up expenditures with tax receipts and bond sales. That’s its cash management function.

But certain people interpret government spending as if gross flows spent by government are done in a naturally limitless way using the medium of exchange that government itself can create – in effect, in the context of your definition - without considering the actual netting effect that you’ve focused on and which government actually adheres to in its own cash management.

That allows those people to say things with a somewhat hyperbolic flair, such as “government spends by crediting bank accounts …” etc. etc.

You may get my drift.

But say no more.

[Nick Rowe]

JKH: I think I get your drift. But I think what those people are doing is simply consolidating the government's and central bank's balance sheets into one. So that government liabilities to the central bank (the government bonds owned by the central bank), and central bank liabilities to the government (the government's chequing account at the central bank) disappear. The central bank itself disappears. Taxes are then paid in government money, which is burned. And bonds are sold to the public for government money, which is burned. And the government prints money, spends some of it on goods, and spends some of it on government bonds, which are burned. G + rB = T + deltaM + deltaB.

[JKH]

Right - although they're consolidating more than the balance sheets. They're consolidating operations, and in doing so implying medium of exchange patterns in the flow of funds between government and the private sector that currently don't exist.

[Nick Rowe]

JKH: I think that's right. But I think it's OK, in some circumstances, as a simplification. But you have to keep your head straight when you translate it back into the real world, where only the central bank can create money, and the rest of the government is no different from anyone else, except it can force people to give it money (taxes), and owns shares in the central bank, and has a chequing account at the central bank (though it lets the central bank control that account, and transfer money between it and the other accounts).

[JKH]

I agree

"keeping your head straight" as in maintaining a balance between the two perspectives for the information and insight they can each provide

but the mere consolidation of balance sheets doesn't change anything; balance sheets are consolidated all the time for actual deconsolidated conglomerate operations

in this case, it can illustrate the effective end point for what might be two quite different operational routes of getting there

one of those routes is factual - just slap together the balance sheets of what happens today, and you can see the effective net position

the other route is counterfactual, assuming it is different from the first route, and if it is not, there really is no point to the exercise other than to identify the consolidated end point, which is a pretty standard thing to do

that recent Krugman post on consolidation was pretty good and clear

he was talking about two different routes to get to the same end point, more or less - ex ante versus ex post QE in a way of thinking

I think that different route idea is the clear way of thinking about it

[Bill Woolsey]

Nick:

This has to be wrong, or rather wrong headed.

There must be a framing problem.

Are the balances people hold in their checkable deposits money?

There is no change in these deposits or how they are spent or received.

But the clearinghouse makes a change in its settlement rules.

The character of the deposit accounts change. But they are just the same from the point of view of these holding them, spending them, and receiving them.

Now, back to the clearinghouse.

Rather than thinking about frequency of meetings, just assume that payments are processed continuously.

I think what you call continuous settlement amounts to is a zero overdraft rule.

There may be payments in line to be processed right now, but if the balance hits zero, it bounces.

Another rules would be to require that overdrafts be covered by the end of the day.

Anyway, you can see that tightening up on overdrafts at the clearinghouse requires that banks keep larger balances at the clearinghouse.

Consider the following rule:

Payments are processed continuously and banks have net credit and net debit balances. Interest is charged on debit balances and paid on credit balances. Net debt balances must be collateralized. The securities used for collateral are never directly transacted for goods and services.

Net balances always sum to zero. By raising the gap between the interest rate the clearinghouse charges and pays, the equilibrium desired balances can all be zero. Basically, it motivates an interbank lending market and motivates lots of transactions to keep clearinghouse balances, positive and negative, low.

Now, suppose the interest rates paid and charged are brought closer together, so that some banks desire to hold positive balances (and other banks negative balances.)

Suppose the clearinghouse buys up the collateral. It provides all of the banks with mutual fund share balances and tightens up the overdraft rules.

Checkable deposits used by the nonbanking public change their character?

The clearinghouse's sole asset is discount lending. Banks with net debit balances borrow, with appropriate collateral. Banks with credit balances hold balances at the clearinghouse. Overdraft rules? Well, banks that would otherwise have an overdraft get a discount loan.

Does that change the nature of the checkable deposits?

Now, if we are not worried about the medium of exchange, but rather medium of account, there is an entirely different analysis.

[Nick Edmonds]

Nick and JKH,

What in your view is the medium of exchange for payments between banks and their customers, e.g. what is the MOE when my bank credits my deposit account with interest?

[JKH]

Nick,

From an accounting perspective, that example is a debit to equity and a credit to deposits. It's entirely internal.

So not sure how that particular example falls into medium of exchange.

In general, I think the medium of exchange for bank customers is deposits (as opposed to central bank reserve balances) and currency.

So that example could be deposits as MOE I suppose.

But I'm a bit lost as to how others think about this in general or if there is any standard view on it.

[jt]

For BoC, http://www.bankofcanada.ca/wp-content/uploads/2010/07/lvts_primer_2010.pdf & http://www.bankofcanada.ca/wp-content/uploads/2010/06/dsouza1.pdf
At the margin, settlement balances or BoC overdraft is the MOE, for LVTS Tranche 1. For Tranche 2, you could say the collateral (AAA bonds; mostly gov) is the MOE.

[Nick Edmonds]

JKH,

It seems odd to me to interpret this as deposits being the MOE, as there doesn't really seem to be a sense in which a deposit has changed hands. We could perhaps say that what has really happened is that the bank has created a deposit in its own name and then transferred it to me. But this seems to me to be making things unnecessarily complicated. But then what is the alternative if we want to have an MOE?

The thing is, I don't think that having an answer to what the MOE is tells you anything more than you already know from understanding how the payments occur. And in fact it just confuses things. We should be able to deal with questions about how the quantities of checkable deposits, reserves and so on matter without resorting to concepts like MOE.

[JKH]

Nick E.,

I agree with all that

MOE is a semantics/classification exercise that abstracts where the value of abstraction seems questionable - maybe it should even be kept deliberately vague and flexible - the operational detail is the real thing

[Odie]

Nick,

"The velocity of circulation is undefined in that example. Even if you assume that people use currency for all transactions, that tells you nothing about the average stocks of money held, and that they want to hold."

Does that not touch on the Baumol-Tobin model of money demand for transactions?

"It is the three women's ownership of a clearing house, rather than their ownership of banks, that makes their IOUs function as a medium of exchange."

Is a bank not also a clearing house? It settles the loans it has issued with deposits of its customers.

[Odie]

JKH, Nick E.,

Thank you. I was always wondering whether I overlooked something important in those discussions about MOE/MOA etc. I just did not see the point in making those distinctions.

As a side note: I once read in an article that large commercial transactions (such a company takeovers) are paid for in T-bills (which then would act as MOE I assume). Makes kind of sense considering that foregoing interest even for just a day adds up when you talk about a billion $ transaction.

[Roger Sparks]

Nick R.,

The frequency of clearing-house-meetings becomes meaningless if account balances are sequentially maintained. In other words, if the clearing house function is only to debit and credit accounts, never to debit nor credit unless funds preexist, then the frequency of clearing-house-meetings does not matter.

If the clearing house never meets, re-balancing would never occur.

[Nick Edmonds]

Odie,

I don't think it's generally the case that large commercial transactions are settled in T-bills and I can't see why that would need to be done as a payments matter (although there may be tax or legal reasons in certain specific cases). Such transactions are usually set up so that all the payments (including loan drawdown or asset liquidation) happen simultaneously. No-one has to worry about lost interest, unless they're left with an open position at the end of the day, which shouldn't happen if it has been set up properly. But it still raises similar issues, because often the "money" used for the payments does not exist at the beginning nor the end of the day and the ambiguity over order of payments makes it questionable whether it ever really exists at all.

[Majromax]

I think that your post combines two (and a half) different but equally important mechanisms by which the IOUs become money.

The first (and a half) mechanism is by which IOUs can be exchanged and cancelled -- the IOUs become bearer instruments. That's the means by which *particular* IOUs can become media of exchange; otherwise Betty would keep Alice's IOU and then write a new one to Cathy, who keeps Betty's IOU and then writes one herself to Alice. That would create a situation somewhere between #2 and #3, and it would leave intact liquidity risk -- a single $20 bill would suffice to clear the entire exchange of IOUs, but without that then the situation is in mutual default. (This also creates nasty ordering problems for just how balances have to be resolved.)

For a visual metaphor, exchange and cancellation of IOUs replaces a pair of one-way "pipes" between people (pairwise) for outgoing and incoming debts with a single pipe, which can flow either way depending on net debt.

The clearing house goes one step further and allows IOUs to be equivalently exchangeable regardless of the originator, and it's the mechanism by which a debt to Alice can be cleared by a credit from Cathy. In the visual metaphor, it replaces pairwise pipes with one per actor to the central reservoir. That's by far the bigger reduction in financial tubing (going from 2 * 2^N pipes to 2^N with bearer-IOUs to N with the exchange), but it first requires IOUs be bearer instruments.

> The extent to which chequable demand deposits are media of exchange depends on the percentage of transactions that net out in the clearing house, which is a negative function of how frequently the clearing house meets.

It's not quite a time-frequency as much as permissible margin levels. The circle of on-demand IOUs is ultimately backstopped by the participants' cash reserves. "Full reserve checquing" could be maintained by forbidding any participant from ever writing a cheque for more than her cash balance net of outstanding cheques at any moment, regardless of yet-uncleared deposits. Then the clearing house could settle transactions in full or in part, at any arbitrary moment of time, without relying on cheque-cancelling.

Allowing partial reserve chequing does make cheques a medium of exchange in their own right, as clearing the resulting web would in general require the cancellation of ofsetting debts. Settlement for cash would still be possible (with any amount of hard currency as a backstop), but would require a preferential order.

To clear up my rambling a bit, the critical factor in cheques-as-money is whether Alice can write a check (to be paid "on demand") on the basis of a cheque-in-hand from Betty.

[Vincent Cate]

Is it reasonable to view all the other definitions of money supply beyond base money as just ways of moving base money ownership around faster?

I have a post looking at the equation of exchange and base money. From this it really looks like high inflation is coming but if you look at other definitions of the money supply it is not so clear we will have inflation. I seems like this post sort of addresses the issue.

http://howfiatdies.blogspot.com/2014/01/how-we-know-inflation-is-coming.html

[Nick Edmonds]

Comment in spam?

[I just checked - I can't see it. Sometimes they just disappear... SG]

[Jason]

Is a recession then a change in the frequency of clearing house meetings (due to e.g. less trust among the agents)?

An increase in the amount of currency should then offset the drag (recession) and a decrease in the amount of currency should create drag (cooling an overheating economy).

This view would lend some credence to the idea that the financial crisis caused the recession in the US -- everyone meeting up to clear their CDO's and other derivatives. (The Fed could have offset it by issuing enough currency to cover the increased rate of clearing, but the proximate cause was the increase rate of clearing).

[JP Koning]

I'm not really sure how we can deploy this classification scheme. We know that the economy's unit of account is determined by central bank base money... the extent to which chequable demand deposits are (or aren't) media of exchange doesn't change this fact. With the price level being set by the central bank via manipulation of the base, what is the macroeconomic significance of the extent to which deposits land in the "media of exchange" category?

[Too Much Fed]

Vincent Cate said: "Is it reasonable to view all the other definitions of money supply beyond base money as just ways of moving base money ownership around faster?"

JP Koning said: "We know that the economy's unit of account is determined by central bank base money... the extent to which chequable demand deposits are (or aren't) media of exchange doesn't change this fact. With the price level being set by the central bank via manipulation of the base, what is the macroeconomic significance of the extent to which deposits land in the "media of exchange" category?"

No, we don't know the economy's unit of account is determined by central bank base money.

Let's say people only use demand deposits (no currency) and there is only one commercial bank with many branches so any central bank reserves do not move.

M*V of currency plus M*V of central bank reserves = NGDP

0*0 plus central bank reserves*0 = NGDP

0 plus 0 = 0 (NGDP)

NGDP won't actually be zero because prices will be set in terms of demand deposits. The demand deposits will also have velocity.

[Too Much Fed]

Nick said: "Suppose Betty is temporarily short of currency, and gives Amanda an IOU for $20 in exchange for the haircut. And then pays off (buys back) her IOU a couple of days later for $20 in currency (maybe plus interest)."

I am going to say the IOU is the bond/the loan part. It looks like your clearinghouse is for bonds and that the bonds all trade at face value. If they are bonds, the prices would most likely vary from face value.

Plus, let's try the scenario like this. Amanda saves $20 in currency. Betty is temporarily short of currency. Betty writes up a new bond (IOU). They asset swap. Betty gets the $20 in currency. Amanda gets the new bond (IOU). Now do a second asset swap. Betty gets the haircut. Amanda gets the $20 in currency back.

[Too Much Fed]

Bill Woolsey said: "Now, if we are not worried about the medium of exchange, but rather medium of account, there is an entirely different analysis."

Why can't currency and demand deposits be both MOE and MOA?

[Tom Brown]

Too Much, you've been asking about MOE, MOA (and UOA?) for a long time. Have you ever read some of the articles on these three subjects over at JP Koning's site? He's got some good ones and he does a great job summarizing a lot of different views presented over the years: Rowe, Sumner, Woolsey, Glasner etc. You can use his search box to look them up (type in "MOE," "MOA," or "UOA"). He also has some interesting historical examples. JP's views evolve slightly on the subject, so make sure you read some of the most recent posts too.

[jt]

@JP Koning | January 03, 2014 at 04:53 PM
"what is the macroeconomic significance of the extent to which deposits land in the "media of exchange" category?"
I guess it depends on whether one thinks that a bank run has macroeconomic significance.
If there is doubt about the IOU than there would be a run on the bank. Perhaps this is why the clearinghouse determines the MOE; it's where the rubber hits the road?

[Too Much Fed]

Tom Brown, here is one.

http://jpkoning.blogspot.com/2012/11/discussions-of-medium-of-account-could.html

comment at December 9, 2012 at 1:13 PM

"The definitions I'm using for medium of account, unit of account, and medium of exchange come from NME (see comment above). Here is an example:

"A medium of exchange is an asset that is widely accepted in trade and to settle financial obligations. Currency notes or transferable bank deposits are typical examples. A medium of account is the commodity defining the unit of account. A unit of account is a specific amount of the medium of account. For example, for the gold standard the medium of account is gold, while the unit of account might be one ounce or one pound of gold of specific purity. A unit of account is the unit in which the medium of exchange and other assets are denominated and in which other values and prices are expressed."

Warren Coats, 1994."

Not sure about this part, "A unit of account is the unit in which the medium of exchange and other assets are denominated and in which other values and prices are expressed."

The total amount of currency is MOA. OK
$1 of currency is UOA. OK
Currency notes and transferable demand deposits are MOE. OK

Make currency and demand deposits 1 to 1 fixed convertible. Start with all currency. Next, add a bank. Someone saves $100 of currency and buys new equity from the bank with a 10% capital requirement. The bank creates $1,000 in new demand deposits. The demand deposits should be accepted just like currency. If not, convert the demand deposits to currency. Overall, $100 in currency saved and $1,000 in demand deposits or currency "dissaved". MOA is increased by either demand deposits or currency.

It is not create more demand deposits and they fall in value relative to currency.

[Philo]

{I hope this comment is not redundant; I haven't read the other comments.}

“Suppose all three women . . . give their IOUs in exchange for services. And all three repay those IOUs with currency a couple of days later. There were still three transactions of $20 for currency. Again, I say that those IOUs are not media of exchange.” Why not say the IOUs *are* media of exchange, since they were exchanged for services? Subsequently there were three *further* transactions: the IOUs were retired, each for $20 of currency ($60 total).

If instead of being retired for currency the IOUs are torn up in a clearinghouse session, you agree that they were media of exchange. So at the time when they were exchanged for services *you don’t know whether to call them media of exchange or not*: you will say ‘no’ if they are retired for currency, ‘yes’ if they are torn up in clearing. I think you should be able to tell whether something is being used as a medium of exchange in a particular transaction *just by examining that transaction itself*; you shouldn’t have to wait for later activity.

One might object that *general acceptance* is implied by the term ‘medium’, but to say that a particular person’s IOU’s has been accepted in exchange in a particular transaction does not imply that it (as well as items like it) is *generally accepted* in exchange for goods and services. But then you should not have counted, e.g., Betty’s IOU as a medium of exchange, even if it was subsequently torn up in clearing: *Amanda* accepted it in exchange for services, but that does not show that it would be *generally accepted*.

As I was leaving Toronto last year to return to the U.S. I had run my stock of Canadian currency down to where I couldn’t pay the cab driver who took me to the airport in Canadian currency. He graciously accepted American currency in exchange for his service. Perhaps he would have accepted Euros or Pounds Sterling or Yen. Should we count these moneys as media of exchange because they were accepted in one, perhaps atypical, transaction, or would we require that they be generally acceptable for transactions in Toronto (in Ontario? in Canada?).

[Squeeky Wheel]

Great to see JKH chime in here.

Nick,
I've done some technical consulting on a clearing house system, and can corroborate most of what you've said. Gross settlements (large volumes between banks) are generally done real time and involve movement of reserves - just as you say in the comments. Some clearing houses use "pre-paid" balances - must deposit to the clearing house and maintain positive balance; some use "post-paid" balances - must stay within some limits at all times and make deposits within some timeframe to ensure some normal regulated levels.

Consumer clearing in many countries is daily. Here netting is done over the course of a day (multilateral as JKH describes) with net balances of consumer transactions sent to the real-time gross settlement system daily. Because the clearing/authenticating/verifying of transactions is done daily, consumers get lousy inter-bank service (>= 1 day to transfer). Here the banks' acceptance of payment and settlement are chronologically linked.

The UK (and India iirc, and soon several more) have fixed this tragedy with real-time clearance. In this case, the systems authenticate/verify and clear the transaction in real-time (<30 seconds). The banks accept payment and credit/debit the consumer accounts in real-time. The net balance for consumer transactions is amended, but not yet settled. A net balance for consumer transactions is maintained throughout the day. The balances are still settled through the gross system once a day. Here the acceptance and settlement in reserves are chronologically de-linked (to a limited extent).

[Nick Rowe]

Bill: "This has to be wrong, or rather wrong headed."

Well, it's certainly a bit weird. My result surprised me!

Here's the basic idea:

Suppose we start (as I did) with an economy where there are $60 worth of transactions of services for currency every period.

If we introduce IOUs, and we still get $60 transactions in currency every period, those new IOUs do not replace currency. So I want to say those IOUs are not MOE.

If we introduce IOUs, and we now get $0 transactions in currency every period, those new IOUs do replace currency. So I want to say those IOUs are MOE.

If reserves are like currency, then having an overdraft with the central bank for $20 is like borrowing $20 currency from the central bank.

I'm not sure about any of this, Bill. Still thinking.

[Gizzard]

One thing I'll point out Nick.

It seems your first example using just currency kind of negates the QTM view of inflation monetarist a are so fond of. The price of anything could never be more than 20$ as that is the only income being received, regardless of how fast it circulates. At any one time no one could charge more than 20$ as that's all there is so velocity is irrelevant. Yes GDP can rise but not prices. IOW working more will raise GDP but not charging more

[Nick Rowe]

Gizzard: Nope.

Amanda's income is always Betty's expenditure. But Betty's expenditure is not always equal to Betty's income. Betty can spend more than her income, by spending down her stock of currency. Or spend less than her income, by increasing her stock of currency. And even if the total stock of currency is fixed, nothing prevents all of them spending it more quickly, or more slowly.

[Steve Roth]

Late as usual here...

@JKH: "I'm a bit lost as to how others think about this in general or if there is any standard view on it."

Likewise. And that, I think, speaks volumes about the central terminology here, "Medium of Exchange," and the failure of economics to work with a coherent or standard definition of "money."

That lack of an accepted and coherent definition, hence understanding, explains to me why Nick had to write this post, and why JKH is at a loss (as am I) to understand if there is a "standard view," and what it is. Rather like discussing physics without a coherent and standard, accepted definition/understanding of energy.

I much prefer "units of exchange," which for me is synonymous with "financial assets." ("Medium" here is conceptually intractable, at least for me.) Dollar bills (which are just physical tokens representing account-book tallies), reserve balances, bonds, CDOs, whatever.

All financial assets embody money, which for me is exchange value that cannot be consumed by humans. (Apples have both consumption/"use" value and exchange value. But nobody would call apples "money.")

Money is (unconsumable) exchange value embodied in financial assets. It doesn't exist absent that embodiment.

If we start talking about reserves and dollar bills as units of exchange (financial assets embodying exchange value aka money), with the value designated in the unit of account -- The Dollar (not "dollars") -- does it get easier to think coherently about these questions?

A dollar bill or a dollar bank deposit (credit balance) or a dollar Fed deposit: each is a unit of exchange whose value happens to equal one as measured in the unit of account. (Well, until--in the terms used for money market accounts--the "buck gets broken"...)

I think the problem is conceptual -- failing to distinguish, conceptually, between what's always been called "money" (coins, currency, etc.) and what money actually is: exchange value embodied in financial assets.

I have no idea if this helps. But I hope it doesn't seem crazy to suggest that the widespread (if far from standardized or conceptually coherent) terminology, usage, and understanding re: "money" is at the root of all our confusion here.

This is why I complained about Mankiw's intro textbook not starting out with a discussion of money and value. (Sorry, I think price theory is just a decades-long evasion.) Without that crux-ial understanding, economics is like physics before Newton.

[Nick Rowe]

Steve: "All financial assets embody money, which for me is exchange value that cannot be consumed by humans. (Apples have both consumption/"use" value and exchange value. But nobody would call apples "money.")"

I would call apples "money", if people used apples as a medium of exchange and unit of account. People have used cigarettes as money. And cows, and pretty shells, and gold.

Saying that all financial assets are money, and that no good with use-value is money, is a non-starter.

Did Einstein and Newton think of energy the same way?

And see the macro half of Mankiw's intro text.

[gizzard]

Nick Im not sure you got my point (maybe expressed it very poorly....likely). Im assuming there is nothing else other than the 20$ in currency, which I thought was part of your story as well. So eveyones income is simply the other persons spending and everyone spends their income necessarily to obtain the other service.

"And even if the total stock of currency is fixed, nothing prevents all of them spending it more quickly, or more slowly."

Sure there is something that prevents spending it slower, the price of the good. Its 20$, so they can't spend ten and buy anything.

I do think you are on to something about clearing houses and IOUs. IOUs necessarily must have a hierarchy, some must be less likely to default than others in order for them to become more money like.... more valuable.

[Nick Rowe]

gizzard: Ah! OK. I missed that. I wasn't assuming there was just one $20 note in the economy. If I did assume that, it would be hard to price a haircut at $40, unless you got half your hair cut at once, or paid in two installments.

[Too Much Fed]

Steve Roth, what do you think are the unit of account (UOA), medium of account (MOA), and medium of exchange (MOE) in the USA?

[Matthew]

I think the theory is missing something important, but I'm not totally sure what that is. You're implying that the velocity of money depends only on the frequency with which the clearing house meets. Yet in reality we do observe variations in the velocity of money that are positively correlated with the interest rate. A plausible extension of this model that would fit the stylized facts would be that they base their decision on how often the clearing house should meet on the interest rate--at higher interest rates they would want to meet more frequently to minimize the interest costs. This all reminds me of the old Baumol-Tobin model.

[Tom Brown]

Too Much: based on my previous understanding of JP Koning's evaluation of the blog literature on this, and supposing for a moment we are back on a gold standard with direct convertibility at say $30 / oz I would say:

UOA = The dollar. Literally the word "dollar" or the symbol "$". Defined as being equal in value to 1/30 oz of gold.

MOA = gold

MOE = $ denominated reserve notes, bank deposits, coins, etc.

Then at one point in time JP tried to extend this to fiat money with a CB targeting inflation. I think at one time his definitions would thus have been like this (MOE doesn't change):

U0A = "$" defined as being equal in value to a certain sized slice of the CPI "basket of goods"

MOA = the CPI basket of goods (?)

I **think** that was the case (if I recall correctly) at one time. However, I think he's changed his mind since then. He almost had Nick convinced I think too.

Now Scott Sumner would say that

MOA = base money = cash & reserves = a kind of "paper gold"

I'm not sure what he says about the UOA, but probably:

UOA = the word "dollar" or the symbol "$"

and leave it at that. I've forgotten Nick's take. Bill Woolsey was somewhere close to JP previously I think.

So perhaps "Scott's" definition has the advantage of not being tied to a CB targeting the CPI (or tied to the CB's success in that targeting). Scott's definition is divorced from any concrete commodity or good though, and Scott does not agree with Mike Sproul's backing theory.

Since Scott isn't here to defend himself from my sloppy reckless slander, Let's just call that "Tom's" definition:

UOA = $
MOA = base money
MOE = broad money = base money + banks deposits, etc.

How does that work for you?

[Too Much Fed]

Tom Brown, use a gold standard, but gold is not MOE.

MOA = gold, UOA = certain amount of gold like 1/30 of an ounce of gold. Now define convertibility with currency and keep it fixed (with emphasis on the fixed part). Now currency is both MOA and MOE. Now fix demand deposits and currency as convertible 1 to 1.

MOA = gold, currency, and demand deposits
UOA = 1/30 of an ounce of gold, $1 of currency, and $1 of demand deposits
MOE = currency and demand deposits

Now use a "no commodity" standard.

MOA = currency, UOA = $1 of currency. Now fix demand deposits and currency as convertible 1 to 1.

MOA = currency and demand deposits
UOA = $1 of currency and $1 of demand deposits
MOE = currency and demand deposits

[Too Much Fed]

zyxzyxooxwo

[Tim]

The Canadian Payment Association, a separate and distinct organization from the Bank of Canada could probably be considered a modern day Canadian clearinghouse. Prior to the 1980s and 1990s much of the function of the CPA resided in the Canadian Bankers Association which going back to the pre Bank of Canada era played the historical role of a clearinghouse described previously. However, in the 1980s and into the 90s there was a desire on the part of the Bank of Canada and the Department of Finance to split payment system functions into a separate organization that would representing "non banks" like ATB Financial and Desjardins in addition to chartered banks.

The CPA runs two different parallel settlement system one ACSS created in the 1980s to replace a previous CBA ran system handles non critical retail payments on an overnight basis(Additionally ACSS handles check clearing however out of region checks can take more than one day to clear). The second LVTS was created in the 1990s to replace another previous CBA ran system handles critical payments instantaneously.

The Bank of Canada is both a member of ACSS and LVTS in addition to being a member of the Canadian Payments Association.

There are also additional Canadian specific clearing organizations such as Interac(which uses ACSS) and the CDCC.