Wednesday, September 24, 2014

Peter Martin — Loan repayments destroy credit money. Right? Wrong. They don’t.

It is important to distinguish between the IOU of the borrower held by the bank when the loan is issued, which is indeed destroyed on repayment of the loan and the credit money issued by the bank, which is not destroyed.
Modern Monetary Theory: Real Economics
Loan repayments destroy credit money. Right? Wrong. They don’t.
Peter Martin

62 comments:

Dan Lynch said...

I'll have to think about that.

Detroit Dan said...

I think Martin is wrong. The example he gives, where the money is paid back in cash, is the exception rather than the rule and he does not go through the accounting.

A more straightforward example is that the car loan is paid back by reducing the checking account balance of the borrower, canceling the formerly outstanding loan. The money supply outside the banking system has been reduced by the repayment, while the money supply inside the banking system is unchanged by both the initial loan and its repayment.

Tom Hickey said...

In endogenous money, bank reserve balances as assets of nongovernment are inside the system. The loan is an asset in the banks books, and the deposit that cancels the loan is a liability. They the deposit is no longer a liability of the bank, the loan is no longer an asset and the cash or rb transfer is an asset. That is to say there is no decrease in the monetary base that is owned by non-government.

However, when treasury spends the monetary base increases with a Treasury credit to nongovernment and the monetary base decreases with taxation by decreasing the monetary base as an asset of nongovernment. So spending injects net financial asset, and taxation withdraws net financial assets, whereas in nongovernment financial assets net to zero by accounting identity.

Bond issuance doesn't increase nongovernment net financial assets but merely changes their composition without altering the amount.

So spending through government issuance creates reserve balances owned by nongovernment and taxation decreases net financial assets owned by nongovernment. MMT calls this creation and destruction of money.

Bank credit netting to zero doesn't create and destroy money in this sense.

However, bank money is a liability of banks and cancellation of that liability when a loan is paid off might be said to reduce money since it is reduces M1. However, cash and rb used to make the deposits to pay off loans remain in nongovernment as an asset of the banks.

So loan repayment that cancels loans and deposits reduces credit money (M1) but doesn't change the amount of reserves(rb plus vault cash) held by nongovernment in the banking system as nongovernment ownership of the monetary base. The only way to reduce the monetary base is to transfer rb/cash to government thereby reducing aggregate nongovernment NFA.

Detroit Dan said...

So Martin is indeed wrong, according to Tom. As Tom said,

"So loan repayment that cancels loans and deposits reduces credit money (M1) but doesn't change the amount of reserves."

Tom Hickey said...

Not necessarily "wrong." It depends on how one frames it conceptually. I described what actually happens in the accounting. What I omitted was the case in which intrabank netting cancels the loan and no cash or rb are involved so the deposit and loan just disappear as entries. The bank's assets (loan account) are reduced and the bank's liabilities (customer deposit account) are also reduced. No bank asset account gets marked up on the process. M1 decreases. If one chooses to call that "destroying money" I guess that's OK. Same with the withdrawal of rb through taxation. It's OK to call that "destroying money" since it reduces MB and M1. But that is pretty loose terminology, IMHO, and it's clearer just to show what happens operationally.

Detroit Dan said...

That's pretty tortured, Tom. You posted something entitled, "Loan repayments destroy credit money. Right? Wrong. They don't." Then you follow up with seven paragraphs of too much information concluding clearly that "loan repayment that cancels loans and deposits reduces credit money (M1)".

Ralph Musgrave said...

Obviously money is not destroyed when Martin’s £10k loan is repaid because as part of the repayment process, he introduces £10k to the economy from nowhere, or to be more exact, from the central bank. That’s the £10k of base money used to repay the loan.

Why not introduce £20k of base money from nowhere? Then one could claim that when a £10k loan is repaid, the money supply RISES by £10k!!!

Calgacus said...

Yes, Detroit Dan, Martin is not really right, while your example of course is. Keen, if not confused himself, is very confusing here. (From Keen's story, I don't want to bank at those accountants' bank!) For there are two acts of £10k bank credit creation in Martin's story, not one.

Of course the one £10k loan can't cancel both of them! But it does cancel one of them, so loan repayment does destroy bank money. The first bank credit is from the lending, yes, and goes to the car dealer in the end. But the second comes from the deposit of the pound notes. The bank gives the depositor an account deposit = "bank notes" = bank money in return for the BoE state money notes. This bank money is cancelled against the bank loan, not the original loan credit.

I don't know what Keen means about Keynes, who surely understood this. Of course bank money "circulates" - that is what makes bank credit money = negotiable credit. Which is why the borrower went to the bank in the first place, to get an IOU which is more widely accepted than his own.

Tom Hickey said...

That's pretty tortured, Tom. You posted something entitled, "Loan repayments destroy credit money. Right? Wrong. They don't." Then you follow up with seven paragraphs of too much information concluding clearly that "loan repayment that cancels loans and deposits reduces credit money (M1)".

I disagree, DD. I don't like the "destroy" conceit. I don't think it adds anything but a charge. IMHO, it is much clearer to just say what you mean.

MMT holds that taxation "destroys" money. repayment of bank credit is quite different in important respects so to say that both taxation and repayment of loans "destroy money" would be confusing. repayment of a bank loan reduces M1 while taxation reduces M1 and also MB. Banks can add to and subtract from M1, and so can government but only government can add to and subtract from MB. That is to say, banks are involved operationally in increasing and decreasing credit outstanding, but they cannot increase or decrease currency taken to mean cash and reserve balances at the cb. Government has a monopoly on creation and destruction of currency as cash and rb in the payments system.

This relates to the difference between MMT and circuitism. MMT and circutism both hold that money is endogenous. 

MMT also holds that government creates and destroys net financial assets in the private sector fiscally by spending and taxing. The reserves of the central and bank and in the Treasury's account don't count toward bank reserves. So when bank reserve balances are increased when government spends through the Treasury (using the central bank as its agent) the amount of net financial assets in nongovernment increases and when taxes are paid, both deposit accounts and corresponding bank reserve balances are marked down. The decrease in bank accounts withdraws funds from M1 and the transfer of reserve balances of banks to the central bank destroys them. So government injects and withdraws net financial assets exogenously whereas in the endogenous money circuit all financial assets net to zero. Government uses security issuance to drain excess bank reserves created by deficit spending in order to facilitate the central bank's hitting its target in the interbank market if it chooses not to pay interest on reserves and chooses to set a rate above zero.

Circuitists object that this ignores that government is part of the monetary circuit and that the Treasury spends either from revenue from taxes that originate in deposit accounts or from borrowing through the sale of government securities in the private sector that draws down nongovernment accounts. So taxation and public borrowing are done chiefly through entries to deposit account with the banks' reserve accounts mirroring the process in the payments system. Government fund themselves through taxing nongovernment and borrowing from it rather than currency issuance in most modern economies. That is to say, they see a fundamental difference between the Treasury issuing notes/coin and issuing bonds and being required to obtain central bank reserves in order to spend.

Tom Hickey said...

BTW, this is about aggregates so examples involving an individual account may not hit the mark. As Calcagus points out, the original deposit money that a loan creates has its own life in the circuit and is not usually extinguished by the repayment of the loan that generated it.

However, the amount of credit money as a component of M1 is reduced when a deposit is used to repay a loan. Just as the deposit was created out of thin air by a bank crediting a deposit account funded by a loan, so too the deposit that repays the loan (consolidating all the deposits if there is more than one) vanishes into thin air as the bank's deposit liabilities are reduced when the loan asset account and deposit liability accounts cancel each other.

Net financial assets are still zero, but M1 is reduced by the amount of the deposit that repaid the loan. Those funds are no longer on the books in nongovernment.

But if cash or rb were used to settle, the bank has an add to its reserves (bank asset).

State money is fundamentally different from bank money, and there is a money hierarchy.

These are key understandings that not everyone gets.

netbacker said...

Prof Keen responded on twitter
@ProfSteveKeen: Actually @netbacker I have changed my mind on this thanks to Minsky double-entry modeling! Currency issue still relevant though.

Anonymous said...

Private banks create debt and money at the same time. Considering debt to be negative,

bank_money + bank_debt = 0

That equation always holds for bank_money and bank_debt.

Unknown said...

from the bank accountant's perspective, customer deposits represent debts owed by the bank to the depositors, or 'borrowings' in other words. Maybe that's why Keen's accountant friends don't think of marking down deposits as destroying money.

NeilW said...

"That is to say, they see a fundamental difference between the Treasury issuing notes/coin and issuing bonds and being required to obtain central bank reserves in order to spend."

That to me is a failure of linearisation of a circular process.

Any starting point on the circular process is valid, and any process you get from *any* of those starting points is valid. What can actually happen is the logical conjunction of those effects, not the intersection.

MMTs insight comes from operating from the consolidated group accounts of government rather than the individual central bank and treasury.

It is on the group accounts that money appears and disappears, because at that level the government is a bank that buys stuff.

Ryan Harris said...

This all hinges on how you define money.

All credit, private and bank credit alike create "money" and destroy money as it is loaned and paid off.

Forget about the bank and their balance sheets for one second. We all understand how banks mark up accounts. And we "get" that bank assets and liabilities nearly always trade at par with USD because congress insures them and accepts them for payment of taxes.

Think instead about assets and liabilities of non-bank banks. If you (a person) lend me 100k to buy a house, you take the note down to your county court house, record it, you now have an asset, it produces income, it is readily bought and sold in the secondary market. You can now borrow against the mortgage asset. It is "money." And to you it is almost as re-assuring (but not quite) as having a bank reserve. Which is almost as reassuring but not quite as having a dollar bill. Which is almost as reassuring but not quite as owning a treasury bill.

So When I pay off my loan to you, the economy now has one less money asset.

To get there, you have to accept that all financial assets, from accounts-receivables, to mortgages, to my promise to pay you a for my cheeseburger on tuesday, is "money" in the real economy, just like a bank reserve even though it does trade at some discount to a dollar. It is essential to having a realistic view of why in most times, the private sector is pulling or slowing the carriage and why government must always adjust to the needs of the private sector, and not the other way around.

PeterP said...

He is wrong and Keen/MMT are right here.

When you repay a loan you first put the govt money in the bank – bank now has a liability (deposit to you) and an asset. When you then use your deposit to repay the loan the bank’s liability (deposit) is cancelled together with the bank’s asset (loan). So yes, equal number of bank money that had been created by the loan now evaporates. Number of govt money stays constant in this operation, it only changed hands.

Ryan Harris said...
This comment has been removed by the author.
Anonymous said...

This is a reply I posted to Neil Wilson on my blog:

Thanks for your comments. You’ve said “The relationship between a commercial bank and the Bank of England is a simple currency peg. ”

That’s the way I see it too. Commercial bank money is essentially a parallel currency. Its not the same currency though so we have to be careful when looking at the double entries. The two currencies can’t be mixed.

This is the situation prior to the loan repayment. (assets, liabilities)
Commercial Bank Liabilities (Credit Money)
Purchaser (0, 10) Vendor ( 10, 0) Comm Bank (10,10)
Central Bank Liabilities (BoE Money)
Purchaser (10,0) Cent Bank (0,10)

After the loan repayment:
Commercial Bank Liabilities (Credit Money)
Purchaser (0, 0) Vendor ( 10, 0) Comm Bank (0,10)
Central Bank Liabilities (BoE Money)
Purchaser (0,0) Comm Bank (10,0) Cent Bank (0,10)

Note that assets = liabilities in each currency both before and after the repayment.

So the commercial bank has given up its asset in its currency (or own issued liabilities) in exchange for the acquisition of the asset of central bank currency (or their liabilities). It still has its liability towards the vendor, for whom nothing has changed at all.

So I’m sorry that Steve seems to have deserted me but I think I’m still with Keynes on this one!

Dan Kervick said...

It all depends on the way the loan is issued and how it is repaid.

Suppose you have a deposit account at your bank containing $50,000, and you then borrow an additional $10,000. The bank's total deposit liabilities have increased to $60,000. Let's say you spend that additional money by making payments by check and check card to other depositors at the same bank, so that the total $10,000 moves from your account to other people's accounts at that bank. The bank's deposit liabilities have not changed; they have just been redistributed, so that the deposit liability to you is $50,000 again but the liability to other depositors has increased by $10,000 total. Now suppose you repay the loan by drawing on your own deposit account. As a result, the $50,000 liability to you is reduced to $40,000, and thus the bank's total deposit liabilities are reduced back to their orginal level. So $10,000 in deposit liabilities have been "created" and then "destroyed".

But in the case where cash is involved, things might indeed be different. To simplify, forget the loan mechanics and just think of a simple situation in which the same depositor walks into a bank with $10,000 in cash and deposits it. The bank increases the deposit liability to that depositor from $50,000 to $60,000, but in exchange its own cash assets are increased by $10,000. End of story; that's it. So the total amount of money in circulation among the non-bank public hasn't changed, but there are $10,000 more in commercial bank liabilities and $10,000 less of central bank liabilities. On the other hand, the commectial banking sector has $10,000 more in cash assets - which are liabilities of the central bank - offset by a $10,000 additional liability to the non-bank public.

Unknown said...

petermartin, you're complicating things unnecessarily by assuming your 'purchaser' gets £10 BoE notes from somewhere.

Unknown said...

Dan Kervick, in your second example M1 deposit money has still increased by $10,000.

Dan Kervick said...

Yes y, didn't I imply that? But even though M1 deposits have increased, M1 overall has not increased, since some cash perviously held by the public has been moved out of public hands into bank hands.

If a bank issues a loan on Monday and is repaid with cash on Tuesday (for example), then the only difference between that situation and the situation where the depositor simply deposits the $10 K in cash in exchange for a $10 K increase in his deposit balance is that in the first case 24 hours have elapsed between the $10 K balance increase and the $10 K cash handover. Thus, during that 24 hour period the depositor has $10 K from the bank on credit

NeilW said...

It's the balance sheet totals that show up the creation/destruction.

Expanding balance sheet totals are creation and shrinking is destruction.

There is no match between the particular assets and the particular liabilities. Both those sides of the balance sheet have their own lifecycle - until somebody who owns some bank money and owes some bank money bring them both together in a repayment.

At which point the two are annihilated like matter and antimatter.

Tom Hickey said...

NW: "MMTs insight comes from operating from the consolidated group accounts of government rather than the individual central bank and treasury."

Right. This is the fundamental point of contention.

NeilW said...

"Right. This is the fundamental point of contention."

Never understood why. Consolidation is standard accounting practice and an International Financial Reporting Standard.

Specifically to eliminate transactions that are entirely internal and distracting to the overall view of the group.


Unknown said...

"This is the fundamental point of contention"

some people don't seem to get that taking money out of your right pocket and putting it in your left pocket doesn't actually give you any more money.

Dan Lynch said...

Banking is not my area of knowledge/interest but I have a couple of thoughts/questions on the subject.

A PUBLIC BANK: Suppose the federal government makes a loan directly to a citizen (say a student loan). That injects new money into the economy. When the loan is paid off, that money is removed from the economy (plus the interest is "removed", too). I don't think any MMTer's will disagree?

PETER MARTIN's PRIVATE BANK: A private bank makes a loan by creating money out of thin air. Suppose, as Peter Martin claims, the money is not destroyed when the loan is paid off, but instead is cash sitting in the bank's vault. Does this scenario imply that the bank makes nearly 100% profit on the principal alone (never mind the interest) since the only expense to the bank was the cost to maintain reserve and capital requirements, administrative costs, and risk ?

Does the bank report this 100+% profit on its income taxes?

If Peter Martin's claim is true, then banking must be a very lucrative business, and banks should probably be paying more taxes?



Dan Kervick said...

@Dan Lynch:

No, if we leave the interest out of the story, then the bank has not profited. Consider the numbers I used. Suppose you have a deposit account with $50,000 in it and borrow an additional $10,000 on Monday. The bank increases the balance in your account from $50,000 to $60,000 and thus has increased its liability to you by $10,000. Then on Tuesday, you repay the loan with cash. At that point, the net change is that the bank has $10,000 in additional assets in the form of cash, but $10,000 in additional liabilities in the form of balances in depositor accounts. The bank's net equity position hasn't changed; neither has yours.

Tom Hickey said...

What happened is that when the bank extended the loan and the loan was drawn down the bank had to deliver rb to the banks at which the checks were cash or drew down vault cash. Bank reserves are rb and vault cash. So the bank's asset accounts reflecting state money where drawn down. As the loan was repaid either cash or checks drawn on another bank, then that state money is recouped.

If the loan and repayment was entirely intrabank, then netting is operative and no state money is involved.

So the revenue to the bank is the interest and the profit is the revenue less expenses involved. This is determined by the spread, operating expense, and fixed expense.

The bank is essentially being paid a fee for risk management that government delegates to banks for capital allocation. Bank lending essentially liquifies collateral, the collateral offsetting the banks' risk.

This gets the the politics out of capital allocation and delegates risk management to the private sector. Banks are agents of the government and have access to the liquidity of the central bank. Small depositors also have a deposit guarantee that transfers some risk to government to increase financial stability.

Plain vanilla banking is boring and that's the way it should be. Banks, of course, have successful gotten around this so that they can engage in "financial innovation" to increase profits. Those should be kept separate in the interest of financial stability.

Dan Kervick said...

But it is entirely possible for borrowers and other depositors to make payments to one another without bank cash reserves ever being drawn down. Bank deposit balances can move rapidly move from one depositor's account to another depositor's account without any cash ever entering the equation.

However, if there is any central bank money in broad circulation, it got there via previous commercial banking operations.

Detroit Dan said...

"The first bank credit is from the lending, yes, and goes to the car dealer in the end. But the second comes from the deposit of the pound notes. " [Calgacus]

Thanks Calgacus. I was coming to that conclusion somewhat independently, but you helped me get to the right spot. Basically, any deposit of currency creates credit money, and any withdrawal destroys credit money.

Tom Hickey said...

To clarify, state money only comes from the government through the agency of the Treasury/central bank. State money is comprised of bank reserve held by banks, including rb at the cb and vault cash obtained by exchanging rb at the cb. Bank reserves (rb + vault cash) are bank assets and central bank liabilities. This includes rb and cb notes. (Notes and coin issued by the Tresury and Treasury notes are not cb liabilities, but they play a minor role of most modern economies.)

Cash is obtained by exchanging reserve balances at the cb. Cash is held by banks as vault cash (bank asset) to meet the window demand. Cash that passes through the window reduces vault cash and increases cash in circulation. Non-banks have the option of exchanging bank deposits for cash, making bank money created by loans convertible into state money on demand at par. (This blurs the distinction between state money and bank money in the minds of most people, who assume money comes from the government since they associate money chiefly with cash.)

State money only comes from government spending or lending. Generally, it is the Treasury that spends into the economy, which results in the crediting of banks' reserve accounts at the Fed (rb = settlement balances) with banks' crediting customer deposit accounts. So rb is a bank asset and a cb liability, and a government deposit is a bank liability (deposit account credit) and a customer asset. Government lending occurs through the central bank in the interbank system, and that lending is only to banks.

In modern banking, the Treasury issues bonds that are exchanged for banks' rb, and the cb exchanges bonds for rb. This exchange of bonds for rb and vice versa by the cb just changes the composition of state money from a non-zero instrument to a zero instrument or vice versa, leaving the amount unaltered (excepting interest paid). So MMT considers the monetary base to include both rb and bonds, since these are swapped around in the course of normal operations that target the policy rate. Similarly, government bond markets are highly liquid and nongovernment can easily swap bank deposits for bonds and vice versa. Short term bills are generally considered to be equivalents. For example, institutions hold "cash" in the form the T-bills.
(continued)

Tom Hickey said...

(continuation)
The only state money that is used by the public is cash. Bank reserve balances stay in the payments system as a bank asset and cb liability used for settlement. Functionally, rb are more correctly called settlement balances and they are only used in the payments system, to which only member institutions have access.

Bank deposits are options to deliver cash on demand, but in a modern economy most transactions take place using accounting entries on the books of the respective banks, and rb are only used for settlement after netting intrabank and interbank netting.

So money created by banks (loans create deposits) is different in kind from state money and necessarily nets to zero in the banking system as an accounting identity. When a deposit is created bank liability so is a bank asset in the form of a loan. When the loan is paid down the loan (bank asset) and deposit (bank liability) are extinguished. Transfer of bank deposit among customer accounts just moves existing bank liabilities around in the banking system, while some enters the economy as cash in circulation by withdrawal of vault cash.

This is called the monetary circuit. There is an aspect of this circuit that is exogenous to non-government and it is comprised of government agencies, although the chief focus from a monetary standpoint is the Treasury and central bank, through which government fiscal and monetary policy are conducted operationally. The banking sector, the financial sector, and the public —households and non-financial institutions — comprise the endogenous, nongovernment aspect of the monetary circuit.

The interface between the exogenous, government aspect and the endogenous, nongovernment aspect is through the operations of the central bank. State money is called outside money, and bank money is called inside money. All of the books balance across the various sectors of the circuit. This is the basis of accounting identities.

Calgacus said...

You are welcome Detroit Dan.

I do not think there is a genuine difference between circuitists and MMTers. One difference is that some circuitists did not oppose the Euro - thereby implicitly repudiating their own theories. Some did however, like Alain Parguez, and they are MMTers to all intents and purposes.

Consolidating the Treasury and the Central Bank is not an axiom of MMT, but a theorem of MMT. The point is to do the accounting carefully and consistently. If you do, and feed in the same facts about the real world, then you will come to the same conclusions, consolidate or not. After consolidating, you will see less creation and destruction of credit, because you have eliminated Treasury/Central Bank transactions.

Bank deposits, bank money are not fundamentally different from state credit, state money, and loan repayment is not fundamentally different from taxation. The same word "reflux" can be and is used for these processes. The only difference is that bank money is a bank debt, a relationship with a bank, while state money is a state debt, a relationship with a state. The difference between bank money & state money is not mystical, and is no more than the difference between the UK's pound state money and the US's dollar state money. The virtue of MMT is that it clearly treats bank "credit money" and state money as fundamentally the same, as both being "credit money."

Detroit Dan said...

Well said, again, Calgacus.

I'm also a fan of the concept of the Pyramid of Liabilities as described by Wray: http://neweconomicperspectives.org/2011/09/mmp-blog-15-clearing-and-pyramid-of.html

Anonymous said...

Dan Lynch,

Banking might be lucrative but not because banks can create money out of thin air. As Minsky said anyone can do that.

Commercial bank money consists of the IOUs written by those banks. State money consists of the IOUs of government or the Fed in the USA.

Commercial bank money is destroyed whenever the payee asks for it to be converted into State money. The commercial banks have to guarantee they will do that otherwise they'd be in default.

Booms and slumps are caused by either too much or too little commercial bank money being spent. So typically there would be a slump a few years after a boom. Not because borrowers are repaying loans. There hasn't been time for that. But because the taxman demands taxes be paid in State money.

In a boom the government budget figures look good , in neo-liberal terms. That's because the banks create credit money which stimulates the economy leading to increased tax receipts in State money.

Calgacus said...

Thanks again, Dan. Yes, it is an important concept. But state money being on top of the pyramid, the king of the hill is only a behavioral fact of life, not a logical necessity. A few centuries ago in Europe, it was not always true. Some bank money was better than state money. Just the same way that the US dollar is now the primary world reserve currency, while the UK pound was and isn't any more. So each state now has its own pyramid. But it can be useful to draw another pyramid with the US dollar and Uncle Sam on top of even the other state's pyramids. The higher on a pyramid, the more widely accepted a type of money or currency is.

But I am not a fan of "inconvertibility/convertibility" as expressed it in that blog. I prefer when he says things like there is no such thing as fiat money, there is only sovereign money. For the US dollar is and always was convertible, into very valuable merchandise. The US dollar's superconvertibility is what causes it to not be converted so much in practice, because it is so valuable. Bad money drives out good, which is treasured and hoarded.

It's like when you visit a supermegaultramall. You see a shop in the front selling the cheapest things, souvenirs and T-shirts proclaiming you visited the Washington Mall. You learn that this dinky shop is in hock to all the other merchants in the mall, and you think the shop owner Sam is a two-bit hustler, a pawn of all the other merchants. But noooo, he is in fact the merchant prince. All the other merchants have to be his creditor to survive, because what he is selling is not just T-shirts, but the right to be in the mall at all, rights denominated in terms of his convertible and sought-after credit, because he, Uncle Sam, owns the mall.

Tom Hickey said...

But I am not a fan of "inconvertibility/convertibility" as expressed it in that blog. I prefer when he says things like there is no such thing as fiat money, there is only sovereign money.

Another point of difference between the circuitist and MMT. They claim that there is no difference between bank money and state money and it's all really bank money, since bank money greatly predominates. According to MMT, it's all sovereign money and banks are agents of the state.

Anonymous said...

Let’s make it all as simple as possible.

Let’s consider the example of someone walking into the Royal Bank of Scotland to borrow one of their £10 notes. The RBS are one of those banks ‘north of the border’ who are allowed to print their own banknotes. They are essentially their IOUs . So these notes are the type of credit money we have been talking about. They tend to be well accepted in Scotland but less so in England.

So they borrow £10 and spend it on whatever! When the time comes to repay the RBS the borrower could , leaving aside the complications of fees and interest payments, give them a Bank of England £10 note, the BoE being the Central bank of the UK, and the debt is cleared. The credit money is still in circulation. It isn’t destroyed.

Alternatively, they could repay with a RBS £10 note in which case the credit money is destroyed.

I probably should have said “not necessarily” rather than “they don’t” but I think that’s all there is to the argument.

Joe said...

This may not be the proper way to do T acct diagrams, but it looks to me like repayment of the loan most certainly "destroys" the deposit that was created when a loan was issued. The 10k of cash that was deposited in the bank added to both the bank's assets and liabilities. So really, in the end, there is still the 10k deposit in the vendor's account, but that's offset by the 10k cash that was put in the bank.
Any issues with what I have? (I hope the formatting works, if not, paste into notepad)


Bank
Assets | Liabiliies
------------------
1. issue loan ) 10k loan | 10k deposit(borrower) (new 10k of bank credit created)
2. buys car ) 10k loan | 10k deposit(vendor) (deposit transferred to vendor)
3. deposits cash) 10k cash | 10k deposit(vendor) (new asset & new liability created)
10k loan | 10k deposit(borrow)
4. repays loan ) 10k cash | 10k deposit(vendor) (loan asset & 10k deposit destroyed)

Borrower
Assets | Liabiliies
-------------------
1. takes loan ) 10k deposit | 10k loan
2. buys car ) 0 deposit | 10k loan
3. deposits cash) 10k deposit | 10k loan
4. repays loan ) 0 | 0 loan

Vendor
Assets | Liabiliies
-------------------
1. ) 0 deposit |
2. ) 10k deposit |
3. ) 10k deposit |
4. ) 10k deposit |

Joe said...

sorry guys, the formatting looked good when I typed the comment. soo, nevermind I guess

Detroit Dan said...

Peter Martin -- Calgacus addressed your extremely atypical example above. And I think you get the gist of it, as you noted that, "Commercial bank money is destroyed whenever the payee asks for it to be converted into State money."

Similarly, commercial bank money is created whenever the payee deposits State money to a bank account. In your (extremely atypical) example, bank money is created by deposit of state money to a commercial bank account.

As Calgacus said (and you apparently did not comprehend), your example shows 2 instances of commercial bank money creation. The loan repayment cancels one of these (eliminating commercial bank money).

Again, this all an extremely unusual situation you have conjured up, where a loan is taken in commercial bank money and repaid in currency (cash). Don't be an idiot.

Detroit Dan said...

Hey Joe -- Your accounting makes sense I believe.

If a borrower pulls out $10k cash to repay the loan, then that reinjects commercial bank money, via the loan to the bank of $10k. Still, when the loan is repaid, the commercial bank money is decreased.

Again, this is not whatever happens. Why would someone borrow $10k if they have $10k in cash to repay the loan? How many people keep 100 $100 bills lying around for such purposes? Total fantasy...

Anonymous said...

"Why would someone borrow $10k if they have $10k in cash to repay the loan?"

It could be a bridging or payday type short term loan to see someone through who works the black economy. So they'd get most of their income in cash.

It doesn't really matter. The theory should hold up for all possible scenarios. However unlikely they may be.

Joe said...

Peter, I don't see the issue at all. When the loan is repaid, 10k worth of bank liabilities (deposit) are destroyed, along with 10k of bank asset (loan). All transactions net to zero.

When the 10k of cash is deposited, the bank gets 10k of assets (the cash), and 10k of liabilities are created (the deposit).

To say that the original deposit's credit money (created by the loan) still exists and the credit money created from depositing the cash is what is destroyed, has no practical difference. Sure, there is technically a difference, but the numbers add up the same regardless, credit/money is fungible.

Either way, when the loan is repaid 10k of deposit(bank liabilities) are destroyed, as well as 10k of bank assets (the loan).

In the case of cash itself, everything still nets to zero. Whoever had the cash before the borrower has 10k less of cash (-10k), and now the bank has it (+10k). And if the car vendor wanted to withdraw his bank deposit, the bank would give him 10k. Everything works.

Dan Kervick said...

Why is this an issue? Peter is completely right. A borrower can repay a loan with a third party liability accepted by the bank - in this case the liabilities directly issued by the central bank. After that happens, total commercial bank liabilities held by the non-bank private sector have been increased (whether in the form of deposit balances of BoS notes in circulation) and total central bank liabilities held by the non-bank private sector have been decreased.

Is this inconsistent with anything important anyone has said before?

Unknown said...
This comment has been removed by the author.
NeilW said...

"For the US dollar is and always was convertible, into very valuable merchandise."

That's not a conversion. That's an exchange. A very important difference.

Conversion means that the quantity of one element is reduced, and the quantity of the other is increased.

So with a gold standard the fiat money is deleted when the gold is released into circulation from the vault.

No deletion, no conversion.

NeilW said...

"Is this inconsistent with anything important anyone has said before?"

No, but you can read to much into it, because you've gone straight to the contra net position.

When you pay off a loan with state money or for that matter a gold bar, first you sell the state money/gold bar to the bank in return for a deposit.

*Then* you use that deposit to pay off the loan.

The net result is that the assets of the bank changes. Loans down, state money/gold bars up.

So it isn't an atomic transfer. It is a sequence of creation and destruction matches from the underlying primitive transactions.

Dan Kervick said...

"It is a sequence of creation and destruction matches from the underlying primitive transactions."

Sure, Neil,but the net effect is the same as if you had simply paid of the loan with the state money or gold bar, without any intermediate change to your account.

Suppose your account contains a $60,000 balance, of which $10,000 was loaned to you by the bank, and you now want to pay off that $10,000 loan with $10,000 in state-issued paper bills which you already had in your possession. If this is executed by first making a deposit to your account with the bills and then immediately paying off the loan from the deposit account, then your deposit balance temporarily goes up to $70,000 and then immediately back down to $60,000 - and the bank now has $10,000 in additional cash reserves.

On the other hand, if you simply decide to pay off the loan from your deposit account, your balance will be reduced from $60,000 to $50,000, and the bank gets no additional cash.

So there are two possible net effects of the loan repayment. One possibility is for the bank to end up with the same total deposit liability that it had before the repayment, but more cash reserves. Another is for the bank to end up with a lower total deposit liability, but no change to its cash reserves.

Now let's look at all of this from a macro perspective. Let's suppose the total balance in deposit accounts held at commercial banks is $5 trillion, and that the public also holds $2 trillion in government-issued currency notes, while the banks hold $1 trillion in government notes as cash reserves. Now suppose that in Year 1, the banks advance $2 trillion in credit to the public in the form of one-year loans, and that in Year 2, all of those loans are repaid, half with pre-existing deposit balances and half with currency notes. (And to keep it simple, suppose the public withdraws no additional cash following the loans.)

So in Year 1, the total deposit liabilities of the banks go up from $5 trillion to $7 trillion. In Year 2, depositors pay the banks back in the amount of $2 trillion, half by drawing on their existing deposits and half by turning over some of their state money. After all is said and done, total deposit balances are now $6 trillion; total cash held by the public is $1 trillion and total cash held by the banks is $2 trillion.

So the public has $1 trillion less in cash assets offset by $1 trillion more in deposit account assets. The banks have $1 trillion more in deposit account liabilities to the public offset by $1 trillion more in cash assets. The net effect is no different than if the public had simply deposited $1 trillion in cash.

Joe said...

Neil, that's what I was trying to get in my posts. When you deposit state money, that actually creates a deposit (credit money). Then when you repay the loan, it destroys that credit money (the deposit).

To me it sounds like Peter's articles are saying that when you deposit the cash, that it's the cash itself that is your asset, then it is transferred to the bank when you repay the loan, discharging your loan but not destroying any credit money. But really when you deposit the cash, the cash becomes the bank's asset and you receive a deposit in exchange.

Or did I totally misunderstand Peter's articles?

Detroit Dan said...

"Peter is completely right." [Dan Kervick]

Uh, no. He's already admitted he's wrong in 99.99% of the cases where loans are paid back with bank money. And in the other 0.01% case, there's some thing else going on (creation of bank money by a deposit of currency) that cancels out the still existing elimination of bank money in the repayment of the loan.

It's not that complicated...

Dan Kervick said...

He's right about his main point that not all loan repayments result in the net destruction of bank money that was created in making the loans. Rather, in some cases, state money is just transfered from the non-bank sector to the banking sector. Nothing about this should be surprising in a financial system where the public makes use of both bank money and direct state money as money.

Detroit Dan said...

Perhaps instead of "Peter Martin — Loan repayments destroy credit money. Right? Wrong. They don’t.", the post should be retitled, "Peter Martin — Loan repayments destroy credit money for all practical purposes (although, technically, you could say that in extremely rare situations the loan repayment take the place of new credit money creation instead of destroying existing credit money)".

Tom Hickey said...

When a loan is paid down in part or full from currency in circulation, a deposit account is marked up with the deposit, and the bank marks up vault cash. M1 remains the same.

Then is marked down in payment of the loan, with the deposit and loan/interest are extinguished in part or full. M1 decreases by the amount of the extinguished deposit.

The bank's reserves (vault cash + rb) increase, so the bank has increased assets. The monetary base increases by the amount of formerly circulating currency to vault cash.

So MB + and M1 -.

Detroit Dan said...

"When a loan is paid down in part or full from currency in circulation, a deposit account is marked up with the deposit... Then (it) is marked down in payment of the loan" [Tom Hickey]

Right. So "credit money" is marked down (destroyed) when the loan is repaid.

I think we all agree that loan repayments do destroy credit money in general. Are we trying to confuse people?

Tom Hickey said...

In think it is important to see that it is not the cash that pays down the loan. It is the cash deposit with the cash going to vault cash. This means that bank deposit and loan are being cancelled on the bank's books.

The bank deposit created by state money becomes bank money with the state money becoming a bank asset that increases the bank's reserves. Reserves (rb and vault cash) are always state money.

State money is only destroyed when it leaves non-government and returns to the state that created it.

Bank money (deposits) is created by loans and is destroyed by loan payment. A loan creates a deposit and a deposit extinguishes the loan, canceling the deposit.


Reserves as a bank asset are not spendable. Clearly, rb that remains in the payments system is not spendable. But neither is cash in the vault (unless it is stolen).

This may seem quirky to some people but cash in circulation is spendable (M1) and vault cash is not since it is part of the monetary base.

Even among smart people, there is sometimes confusion over this.

Dan Kervick said...

"I think we all agree that loan repayments do destroy credit money in general. Are we trying to confuse people? "

Nobody is confusing anybody. We are just talking about conventional banking operations and procedures, not deep laws of economics. Debts can be always discharged in two ways: either by the debtor turning over some asset to the creditor that the creditor is bound to accept, or by the creditor reducing the amount of its total obligations to the debtor.

Yes, it is perfectly possible when employing the first method to institute some procedure whereby the debtor first hands over the asset for additional credit from the creditor, and then half a nanosecond later releases the credtor from that additonal obligation. Or you can institute a procedure whereby the asset is simply turned over and a portion of the original credit is directly reduced, without the intermediate step of bumping up and then immediately bumping down the credit balance. It's completely arbitrary and the end result is exactly the same.

Nothing important hinges on whether or not loan repayments always destroy bank money.

Dan Kervick said...

"In think it is important to see that it is not the cash that pays down the loan. It is the cash deposit with the cash going to vault cash."

It is not important at all Tom, and it may not even be true in all cases. Whether or not this is the operational procedure in place is a completely arbitrary and conventional decision, and nothing fundamental to the underlying economics or monetary system would change if procedures were in place by which the loan is permitted to be discharged directly with the cash, with no additional credit extended and then reduced. The end result is the same, and the difference between the two systems is economically meaningless.

Tom Hickey said...

It may not be important, Dan, but it is in accounting procedure.

If someone loans someone something, including money, the return of the item or money amount discharges the loan. No books.

But in transactions that are accounted for, various accounts get credited and debited, and it to understand what's happened, all that is needed to consult the books.

A cash deposit becomes a liability of the bank, and the corresponding asset is an increase in vault cash. This is similar to rb being deposited into a bank's reserve account at the cb with a direction to the bank to mark up a customers deposit account in that state money is involved.

When a loan creates a deposit, the bank creates bank money.

Significantly, when cash is used to pay down a loan, the cash deposit doesn't decrease M1, since M1 includes both cash in circulation and bank deposits. The cash is no longer circulating but in the bank's vault, counting toward reserves, and a deposit account is marked up — the bank has borrowed the funds from the depositor and has a liability.

But when the customer directs the bank to use the deposit to pay down the loan, then the loan and deposit are extinguished in full or part, reducing bank deposits and therefore M1.

I much prefer to describe what actually happens than make claims that may be ambiguous, as this thread seems to demonstrate. But I learned something thinking it through. So it can be regarded as my thinking aloud. If anyone sees an error, please alert me.

Tom Hickey said...

It is not important at all Tom, and it may not even be true in all cases. Whether or not this is the operational procedure in place is a completely arbitrary and conventional decision, and nothing fundamental to the underlying economics or monetary system would change if procedures were in place by which the loan is permitted to be discharged directly with the cash, with no additional credit extended and then reduced. The end result is the same, and the difference between the two systems is economically meaningless.

That's true, Dan, but when the accounting is passed over even smart people make mistakes. That's a reason there is accounting in the first place. It keeps everything straight. Obviously this is especially important in the case of banks where funds are being transferred around constantly. With DE accounting there is always a clear paper trail.

But it is also important in economics, where economist often make jumps regarding financial matters, thinking it obvious, and then make mistakes.

Ryan Harris said...

Create narrow, irrelevant divisions to define specific types of money, and assets and then argue for hours about how one type of transaction impacts your narrowly defined, practically irrelevant definitions. Rawr. I am economist, hear me roar.