Wednesday, November 9, 2016

The NGC Model, Banking and the Creation of Money

Nick Rowe is a prolific and creative author to whom we all owe thanks. Today he gives us a model that we can use to create a new vision of the creation of money.

Nick writes
You decide to make a new monetary system from scratch. You give everyone a chequing account on your computer, with an initial balance of 0 units. If Andy buys bananas from Betty and pays her 100 units, Betty now has a positive balance and Andy now has a negative balance.
Nick may disagree but I think his model first presupposes a monetary system and then calls our attention to a nearly complete monetary event. The focus of this post will be on the events that had to occur before Nick's model begins. It is these background-events that are interesting and offer us additional insight into the creation of money.

Establish a Frame of Reference

These background events need a frame of reference to make them plausible, easily described, and comparable to actual macroeconomic events. The framework we will use is the National Gift Certificate Model (NGC) which treats money as an analog of the well known gift-certificate.

The possibility of a NGC model of money creation first appears in a comment to Nick's post:

Using the NGC example, the issuing store prepares to sell a gift certificate by printing the certificate. Then the store has a choice: recognize the increase in inventory and expense of production immediately OR wait until actual sale and then recognize the event. The first choice makes sense if the intended use of the gift-certificates is to pay for goods and services; the second choice makes sense if the intended use is to sell the gift-certificate for money in the future.
In either case, the actual creation of money or NGC is not visible to the usual measuring tools available to the public. The actual issuance of new money is more visible but identical to using old money; issuance is just another exchange of goods and services for money/NGC.

It is obvious that any store offering gift certificates can do some accounting behind the curtains as just described. We need to notice that the store is providing not only accounting services behind the curtains--it is providing banking services. With the issuance of a gift certificate, the store is providing a certificate with future value (like a check written on a bank). The store is also providing storage for future value that will be claimed when the gift certificate is presented for redemption (store products will be available in the future which is comparable to bank money that will be available when a check is written).

A Model of Money Creation

We have not yet created money. An analog with newly created money comes from the observation that a gift certificate can be issued by a store in payment of goods and services received. This event can be compared directly with a barter transaction in which goods are traded for goods. The only difference is that this is a partially delayed transaction: the services are delivered first; payment is made later when the gift certificate is redeemed. The creation of a money certificate is easily compared to the creation of this delayed-payment* gift-certificate.

To see the comparison, we look at the creation of money by a bank when a bank makes a loan. The bank creates a deposit in the borrowers name and will ask the borrower to sign a promise to repay the amount borrowed (and a small charge in repayment for the bank's services). This action creates an increase in the money supply and is widely recognized as a creation of money.

If you are like me, you don't immediately see these two events (the store's certificate and bank's new deposit) as being nearly identical. To see this clearly, we need to go behind the curtain. The bank can perform the identical steps just completed by a store preparing to offer gift-certificates as payment. Whether either store or bank actually does these steps in preparation or not, the important take-away is that observably identical results are obtained.

There are two behind-the-curtain tasks that both banks and gift-certificate issuing stores perform:

1.  Prepare a certificate that will be delivered to the customer. A store will prepare a gift-certificate. A bank will open an account with a positive balance in place. In both cases, before delivery to the customer, the name on the instrument will be the name of the issuing entity (or blank).

2. Make a record that the certificate/account has been backed by an asset. A store will reserve goods/value owned by the store.  A bank will create a bond signed by the bank and promising to repay the amount on deposit.  (Yes, this is a bank lending to itself but no problem, the actions are entirely within the bank, hidden behind a curtain.)

With the instruments in place and funded, a store or bank is ready to pay a bill or service a new customer.

When a bank borrower actually comes through the door, he is asked to sign a loan agreement. If he does, the account is signed over to the borrower. The bank's loan document is now secondly backed by the borrowers signed repayment commitment.

We can see that money has been created ad nihilum. Or has it? When the borrower reduces the deposit by spending money the deposit represents, other banks will receive money.  In the fiat monetary system, the bank transferring funds must transfer money that all banks will accept. There has to be a problem if all banks are lending money created by assigning accounts and backing them with bank guarantees.

The Dilution Problem

There is a problem. The observed money supply rapidly increases which effectively dilutes the base money supply that was first used to begin the repeated-lending-sequence. In the modern fiat monetary system, this expansion is controlled by the central bank.

The control process is simple. The central bank creates a block of money using the bank money creating process previously described. This would be considered as "base money". Base money is then distributed to private banks, usually through the sponsoring government when government pays for goods and services. Once received, private banks are allowed to re-loan the new deposits. Control of the loan process is maintained by the central bank with a requirement that a "tax" be collected on deposits in every private bank. This "tax" drains original funds at a rate proportionate to the "tax rate",  amount, and number of loan events. After several events, the entire original money issue will be returned to the central bank, allowing the central bank to know when the original base money supply has been completely loaned-to-limit. (This sequence is partially described in the Federal Reserve document "Reserve Maintenance Manual").

We have seen how the NGC model can be used to understand the creation of money.

Another Look at Nick's illustration--Motive

We now have the opportunity to examine Nick's illustration from a new perspective. We can see that behind the curtain, some entity must have prepared the green entry that Betty would receive in trade. Some entity would have prepared the red entry that Andy received. Further, Andy would have been given the red entry along with the green entry that he later traded to Betty. The entity behind the curtain can be assumed to have authorized all of this activity.

Why would any entity carry out all this activity? One logical possibility is that Andy thought that bananas were worth more than the 100 units and the added obligation-to-repay that he incurred. At the same time, the sponsoring entity could have thought that Andy's obligation to repay 100 units in the future (together with service charges)  was worth more than the 100 units that the entity would render temporarily unavailable. We can safely believe that Nick's illustration has described a three-way mutually advantageous trade.

Nick inconveniently omitted the events that occurred behind the curtain.

The Value of Money

It does not seem logical to extend the creation of money into the value of money. We can safely believe that Betty willingly traded 100 units for bananas but we don't know how many bananas she traded. We only safely observe that 100 units have been found.


The NGC model provides a powerful analogy to processes found in the broader monetary system. This descriptive path-of-creation for money is yet another example in the use of the model.


Thanks to Nick Rowe, whose cryptic yet tantalizing posts on macroeconomics repeatedly inspire many amateurs and professionals.

*[11/10/2016 update] The combined words "delayed-payment" were inserted to call attention to the important fact that the comparison of money to a gift-certificate depends upon the store FIRST receiving goods or services THEN providing a gift-certificate in acknowledgement. The gift-certificate is evidence of an obligation for the store, only payable by trade within the store. We could think of the certificate as being Delayed Evidence of a Barter Transaction (DEBT).

Tuesday, November 8, 2016

Comment on National Gift Certificates as a Money Analog

Antti Jokinen and I continued to exchange comments for a short time on Nick Rowe's blog. The comments resulted in what I thought were some good guidance for future development. I will re-post the comments here in an effort to consolidate the record.

The comments repeated here follow the original posting of my article "National-Gift-Certificates as an Analog to Money".

Antti begins:
Roger: First, thanks for all the acknowledgements! Very kind of you. I might be repeating myself, but here are some comments on your post:
From the point-of-view of the HOLDER of a NGC, the USBS metaphor works quite well. But from the point-of-view of the ISSUER, I can't make it work. To start with, you should define what kind of money you are talking about: only fiat money, or also "bank money"? Then, you must decide how a NGC can be redeemed: say, by buying goods for sale in any store in the US, or just by paying taxes (latter is what Randall Wray suggests, having in mind fiat money). "(private) Bank money", deposit, can disappear when you use it to buy something from any store in the US. Fiat money doesn't, so in that case you'd only trade the (fiat-money-as-a) NGC with a non-issuer (This begs a question: How can a commercial bank deposit disappear when you buy something from a non-issuer?).
If we only talk about fiat money, and by this we mean "central bank IOUs" (not my terminology), then we face a problem Wray/MMT seem to face, too: If fiat money is a "CB/government IOU", then why does it disappear when a private entity behind the MBSs on Fed's balance sheet makes a mortgage payment? Does the government not only allow its IOUs to be used in tax payments, but in mortgage payments too?
As you see, I see problems on the "issuer side" both when it comes to your NGC interpretation and when it comes to interpretation of money as an IOU. I think those problems are very, very hard to solve if we stay within those frameworks.
Antti (Nov. 3, 04:09 PM);
Thanks for reading my post and commenting.
You will be surprised when I say that all of these issues are easily incorporated into the analogy. The difficult part is to explain in a comprehensive way. I begin with a brief background:
Fiat money is nothing more than circulating paper printed by the government. Period. It is given legitimacy by taking care that a bond is issued at the same time circulating paper (or electronic equivalent in every case) is printed. Therefore, it is possible for the central bank to meet with the Treasure (two people from two departments) and exchange products: The CB delivers currency and the Treasure issues a promise to pay it back. Period.
The duration of this distributed paper will depend upon the tax rate charged. Assume that a tax is charged at each transaction (income tax, sales tax, VAT tax). After each tax, less currency remains in circulation.
If there is no tax on a transaction (such as on the expense side for income-tax-on-business-profits), there is no reduction in outstanding currency. This enables the duration of any issued paper to be VERY long.
Banks do not issue money. They only have the character of increasing the amount in measured circulation. Take gold as the example. If gold is the money in use, it is very difficult to increase the amount of gold in-hand. It is easy to write a gold certificate and lend it (as if it were gold in-hand) without telling the owner of the original gold. If this is done repeatedly, the amount of gold on deposit will remain unchanged but the amount of gold CLAIMED will increase. The role of central bank reserves comes into play here to control this process.
With this background complete (in a very sketchy fashion), we need to deal with each of your gaps from the issuers standpoint. We group concerns:
1. Money disappears that can be identified as originating with a bank. This occurs when a loan is paid away. Until the loan is paid, the money issued can circulate between users including government (both as a user and destroyer of money).
2. Money disappears that can be identified as originating with government. Taxes are the mechanism already described.
3. I don't understand the MBS mortgage question so I will skip that (perhaps to my peril).
4. It remains to tie fiat money to NGCs.
It is easy to see that government can allow everything-mentioned-so-far to occur. Not everyone will agree with me but I think everything described already occurs on a routine basis. The question then, is whether a private store that issued gift certificates could put in place each of these processes and procedures? I think this has happened already, visible and embodied in the form of company stores and company towns. The early development of America had several examples of small communities that were basically owned by one entity. In come cases, the community used company money which was the equivalent of paying bills with gift certificates.
This was not a good deal for the workers. The company had control of the interaction of company currency and the greater currency of the central government.
There is no question that private stores can issue gift certificates. It certainly seems possible for the private use of gift certificates to expand to include borrowing, banking and complete use in trade exchange. This expanded use of gift-certificates creates what I would like to call "A National Gift Certificate Economy", even if is limited in size to be no more than a company town.
I have attempted to tie fiat money to NGCs. Is the analogy making more sense now?
Antti: I should revise one line to introduce the actual issuance of fiat money. Issuance does not occur when the money is created (creation is all within the confines of government). Issuance occurs when government spends the newly created money.
The line "The duration of this distributed paper will depend upon the tax rate charged." would be much more informative if it read " After issuance (by government paying it's obligations), the duration of this newly distributed paper will depend upon the tax rate charged. "
About your definition of "fiat money": You talk only about paper, but we have to include bank reserves, right? All "high-powered money".
Antti (Nov. 4 10:52 AM): Antti writes "What bugs me is how to explain that you get something from the government for your NGC when you use it to pay taxes?".
I can't argue an explanation here. The best I can do is to suggest a philosophy. I would suggest that tax on land (we pay annual property taxes in America) and trying to get something from government in exchange for my NGC are the same. One rational philosophy is that both are a payment of "rent". Using land as the example, the American land owner is more-accurately sub-leasing the ground from the government who is the REAL OWNER. This basic philosophy underlays the entire NGC-USBS framework.
Following this philosophy, we could claim that a tax on each exchange of money is a payment of rent for the privilege of using money. Wild?!
Now I would like to change the focus to fiat money, banks, and government.
As I thought about my previous reply, particularly about the timing of issuance, I began to place more importance on this observation: both banks and government create money in a private fashion, which means "behind closed doors". Let me elaborate: Both banks and in the CB-Treasury-trade create money in a "back-room", low visibility, event where they prepare to issue money.
Using the NGC example, the issuing store prepares to sell a gift certificate by printing the certificate. Then the store has a choice: recognize the increase in inventory and expense of production immediately OR wait until actual sale and then recognize the event. The first choice makes sense if the intended use of the gift-certificates is to pay for goods and services; the second choice makes sense if the intended use is to sell the gift-certificate for money in the future.
In either case, the actual creation of money or NGC is not visible to the usual measuring tools available to the public. The actual issuance of new money is more visible but identical to using old money; issuance is just another exchange of goods and services for money/NGC.
Turning now to central bank reserves, I think this FED maintenance manual is helpful:
I understand the manual to require a reserve deposit at the central bank that increases proportionately to the increase in bank deposits. In other words, it acts like a tax paid in positive money. What is positive money? I think it is money issued by the government but how do you separate it from bank issued money? Well, if positive money is taxed at each reissue/new-loan, the positive money will eventually all be back at the CB. This gives the CB a lot of control.
These comments, while they seem somewhat disjointed, contain some very important insight. Particularly important is the thought that the store issuing a gift-certificate has two choices of when to "book" the process. Which choice is picked would likely depend upon the purpose to which the gift-certificate placed.
Thanks to both Nick and Antti for their roles in advancing the theory of macroeconomics.

Thursday, November 3, 2016

National-Gift-Certificates as an Analog to Money

I will begin this post by thanking Antti Jokinen for engaging in a very thoughtful discussion about the merits of comparing money to a gift certificate. While at first glance, the two instruments seem very different, upon closer inspection and appropriate scaling, the two instruments compare very well. Antti's skepticism coupled (with encouragement) helped bring the parallels and differences into focus.

This post is a response to Antti's comment "The problem I see with your approach is how to explain why we would call a piece of paper a "gift-certificate", when it entitles the holder of it to nothing else than being relieved of his tax obligation (or fines or other charges collected by the government). I do see some sense behind that kind of thinking, but there are many people who don't."

Yes, we need an explanation. We think we know what money is (or do we?) and a gift-certificate seems to have many differences from money.

I first wrote about the similarity of money and gift-certificates in my post "Money is Like a National Gift Certificate". That post was prompted by my observation that a hardware store gift certificate was a lot like money in many ways.

When I walked into that hardware store with my gift-certificate in hand, I wondered why I just didn't go over the the clerk and request she change it into money. It certainly would have been easier than walking around the store, looking for something that I wanted only marginally. Of course, the store wanted the sale, not the possibility that they would retire the certificate and I leave the store with THEIR money. (There might be a clue here)

Paper money is harder to characterize than a gift-certificate but let's look for some DIFFERENCES and characterize them:

1. The size of the acceptability footprint is hugely different. Money is nation wide but a gift-certificate is only store wide; but both have boundaries of acceptance.

2. The denomination is hugely different. Money can be any value if a check, a fixed value if a gift-certificate; but both have EXACT values on the traded instrument.

3. The source of the instruments is hugely different. Money can be created legally only by government, a gift-certificate is created by private entity; but both are created by unique human entities.

4. The range of items available for purchase is HUGELY HUGELY different. Money can buy anything, a gift-certificate can only buy things in the store;  but both can only buy things identified as being for sale.

The ownership of items for sale is hugely, hugely different. The issuer of money does not own the items that might be purchased using money while the issuer of a gift-certificate owns/controls the items that can be traded in exchange; but exchange is observed to occur. (This ownership difference may be related to government's ability to collect tax at each exchange (sales tax or VAT tax).

Now let's look at the many identical features of money and gift-certificates:

Both disappear (retired, or if you prefer, "recycled") when they are received by the issuing entity.
Both can be traded before being presented to the issuing entity.
Both have unlimited durability while (at the same time) both can be lost or destroyed by intentional action.
Both can (or could) be borrowed by either the issuing entity or third parties.
Both are usually created (issued) AFTER the issuer has received some physical good or service.
Both can be created (issued) as a gift without exchange of any physical good or service.

We began this post with the challenge of explaining how money could be considered as being a National-Gift-Certificate (NGC). While the differences between money and a gift-certificate are several, the differences disappear when the differences are scaled and adjusted for character of ownership. We need to scale the gift-certificate up and into a "National-gift certificate". After the scale-up is done, it is reasonable to argue that a National-Gift-Certificate is an analog for money.

In an early comment, Antti suggested that the footprint for a United States monetary system could be described as the "United States as a Big Store (USBS)". This is an excellent description that I have used several times since.

If we think of the United States as a Big Store, we can compare the value of money to the contents of that store. Of course only the items on sale could be purchased but the same condition is a restriction in a store issuing a gift-certificates. The value of USBS money would be dependent upon the ability of money holders to buy freely what ever they wanted and their desire to purchase.

We conclude by opining that we can not call money a "gift-certificate", the differences are too great. We need to call money a "National-gift-certificate" which is a scaled up, souped-up version of the familiar gift-certificate.

Thanks again to Antti. Your participation in this discussion is much appreciated.