How to Create Endogenous Account Money

(Additions done on 3.6.2014 marked with *asterisk*, additions on 22.3.2015 marked with ‘+’) Steve Keen has been for long the foremost promoter of the idea of endogenous money. Many economists supporting post-Keynesian theories and all MMT’ers also support that notion. *Some sources describe the short lived thing that is called here the endogenous money of type one. This includes Bank of England and James Tobin.* One reasonably good is in Scott Fullwiler’s but it just claims that banks do the trick by creating loans by extending their balance sheets. Details are sketchy. Unfortunately we don’t have that many good writings on how to create *endogenous money with long lifetime*. My own earlier writing on money and employment describes in some detail how to create arbitrary amounts of money market instruments and share capital where banks or firms make up circular patterns. The goal of this writing is to present the details of endogenous creation of account money in banks.

The post-Keynesian theory of loan issuance is that the loan is not provided from a pre-existing asset provided by depositors. Instead the bank provides loans by debiting the loan customer’s loan account on asset side and crediting the customer’s current account on the liabilities side. It seems questionable whether such kind of accounting practice is legal and will be approved by auditors.

Many people complain that accounting is confusing and they prefer to see things as cash flows. For endogenous money the related cash flow is an arrow coming from thin air. That’s very much to the point but it explains nothing. Loan processing mostly occurs in the world of accounting. To see what’s possible and sensible the things must be presented in the wonderful world of T-accounts. Double entry accounting is a medieval version of Excel spreadsheets specially intended to accurately describe the two aspects inherent in every movement of money.

Case One: Physical Cash

The standard idea is that the bank business consists of receiving deposits and providing loans out of those means. The bank first receives money and that act expands its balance sheet. Then the bank extends a loan from that money which changes the form of asset that the bank has. Providing the loan is not expected to extend the balance sheet.

First the bank receives a deposit and books it as debit cash, credit savings of customer A:


The bank provides a loan out of that money and pays it out in physical cash. The asset the banks has changes from cash to loan provided:


Then customer B says she won’t like to go around with so much cash. The bank opens a checking account for her and puts the money there:


The special thing is that we expanded the bank’s balance sheet with a new asset (cash) and new liability (the checking account balance). We might have arrived to the same balances by omitting the two last cash bookings:


That approach we, however, think now to be improper. It’s OK to expand the balance sheet with the deposit received but not with the loan provided. Or should we maybe agree that a combination of loan draw-down and keeping that money in the bank is a rightful way to expand the bank balance sheet? +The accounting is, of course correct: The bank has a new asset: the loan to be repaid, and it has recognized a new liability: the checking account to be serviced.+ The stunning aspect is that this seems to open the road to providing unrestrained amount of loans. The same batch of physical money can be extended as loan, received as deposit and lent out again repeatedly. The bank’s available cash will not change in any way while providing loans as long as the customer will not make actual purchases with the money in the checking account. Only when the money in the checking account is actually used then the bank has to pay out that from cash account. The bank must run liquidity gap analysis for various time periods as a preparation and take appropriate action.

We omit interest amounts here. The interest receivable and interest payable accounts would accrue those amounts. The margin between the rates is the major income that the bank lives on. Loan commitments or guarantees given will not immediately change the bank’s liquidity and we are not presenting those cases.

Case Two: Account Money

Companies and private persons alike may have their money on a bank account in some bank. It’s basically the same for banks. They may have accounts in other banks. There may not exist such an asset as our own bank account in our own bank. It’s not possible for a bank to owe money to itself any more as that is possible for persons. (In many languages, including Finnish, a bank’s accounts in other banks are called with the Italian word nostro, or ours. Let’s use it here.) From the bank’s perspective money in other banks appears quite much the same as physical money except that payments go via some specific payment system.

For convenience we utilize a temporary account in our bank as an intermediary. That clearing account must be brought to zero balance on a daily basis by doing some interbank transactions. The account to be debited or credited is initially the clearing account. In the evening all transactions to that account will be booked to a nostro account in another bank. Then we do cash management: That nostro likely has some credit limit but negative balances are expensive. We move money in other banks to avoid such situations. Any extraneous money will be placed in overnight deposits. If we lack funding then we might use an available facility to borrow from other banks. Financial instruments act as natural counterparts for lending.

When our bank provides a loan then normally the receiving customer’s checking or current account will be credited immediately. Let’s say our bank approves a loan withdrawal to a customer’s checking account in our bank. The money is assumed to come from another bank where we have it.


In the evening we might possibly like to inform bank XX of the change and move the money to customer B account. The means to move money from one bank to another is some payment processing system. If this were cash withdrawal then we could not take some coins from the safe and hook them to the customer account. Any more can we use the payment system to hook the money to the account as no payment has yet been done. So the bank ended up creating endogenous account money. An alternative would be to move the loan from the bank’s account in another bank to customer account somewhere outside our bank’s accounting. It would be booked as credit nostro – debit loan. That would be similar to withdrawal of physical cash. For customer relations it’s much better to create the checking account in the bank.

Only when customer B makes actual payments outside of our bank then we can readily process it. Let’s say she pays 500 units to somebody having account in bank XX:


The last bookings above are our closing of temporary daily transfer balances. What we see above is that yes, we are expanding the bank balance sheet when the bank provides a loan. When the loan is consumed that much of pre-existing account money will be needed. *The balance sheet in our bank contracts back. In bank XX it remains as is but assets get moved.* Appropriate means in liquidity management are necessary. Only in extremely rare cases that means arranging a campaign to attract depositors. More normal is to issue financial instruments in interbank markets or to negotiate loan limits with other banks. A normal bank will not plan its accounts in other banks to have money enough to cover maximum commitments. Instead the goal is to have near zero balances on such accounts that provide only very little interest. Let’s call the money related to loan withdrawals as endogenous money of type one. It’s intended to be short lived and therefore total volumes should be small. It is, however, nearly impossible to avoid creating it for loans issued. A side note to this is that in the case of bank runs the lenders appear as one more group of depositors provided they have money that has not yet been consumed.

+When the bank extends a loan to some customer then that loan, with it’s future promised repayment, will appear as a new asset in bank balances. For the counter entry in accounting there are two alternatives:

  1. The bank may move existing money to the customer, say by giving the money to the customer as physical cash. Then it all appears as changing one type of assets to another.
  2. The bank recognizes it’s liability to provide the money to the customer. That occurs by increasing the balance on an account available to the customer. That appears as a new liability. The balance sheet increases. New account money will be generated out of thin air.

I have actually seen the first method used. An agricultural bank on Latvia did not run a branch network keeping cash around the country but they only had local loan officers. When a farmer applied for loan and the request was approved then they provided him with a document entitling the customer to raise the money from their account in another, bigger bank. When the money was taken from the first bank’s account it was equivalent to paying it out as physical cash. No new money was created. Anyway in most practical situations both banks and customers prefer the method two. Customer gets his or her money right there and the bank can keep a close relationship to customer without messing around with other banks.

As to the future destiny of that new money it may disappear soon or live longer. If the account holder takes the money out in cash then this reduces both the balance on her checking account and the amount of cash in the bank. Some part of endogenous money disappears.

If the money is used as payment to somebody in another bank: The first bank takes the money from it’s nostro account in some other bank. A reduction in the amount of account money in the lending bank occurs but the receiving bank has increases on its nostro account and the recipients checking account. Only if money is later taken out as cash will the endogenous account money disappear. Similarly if there is a payment to another account in the same bank then the new money continues it’s existence.

A loan repayment in cash implies that an asset, the loan, will be replaced with another, the cash, and balance size remains. Repayment from customer checking account causes account money to disappear as both asset and liability disappear.+

To do things more substantial the financial system must create account money and share capital in a big way. Next part shows how to do that by issuing financial instruments in circular setups.

Case Three: Creating More Account Money

In normal times it’s always possible to arrange funding for a loan. The question is only whether the margins in interest rates provide sensible income for the bank. Let’s look how to create more money. Just to have more credibility we establish our bank A so it initially has equity of 10 million in cash:


Another bank named B initially won’t own items relevant to us. Then we issue a 12 month CD of 10 million, bank B buys that and places payment on our current account with them:


Endogenous money was created above as the other bank recognized it’s liability to us at an account of their own. We have with money market instruments quite much the same as happened with loan creation. Balance sheet expanded in both banks. *The endogenous money received in bank A has the lifetime of the related instrument.* Had they paid by transferring existing money in other institutions then this would not have happened.

After that we provide a commercial loan of 10 m to an industrial customer XX. We recognize that in our books the usual way:


Then company XX uses the money to buy equipment from company ZZ. They ask payment to go to the account of that company in bank C. Bank A pays it from bank B. The remaining accounts are:


The payment causes the account that bank B has with C to go negative. That can be arranged with credit limits but let’s say that now bank B issues a CD to bank C to cover it. The accounts in B:


The more volatile checking account resides now in bank C. From the viewpoint of bank A this account money has the same lifetime as the related instrument. For bonds that can be tens of years. Let’s call it endogenous account money of type two. The essential advantage that all this activity brings to the financial system is the difference in interest rates of loan to XX and checking account of ZZ

When re-marketable or negotiable financial instruments are issued then they generate more liquidity. The issuer gets the money; the originator of money gets that instrument and can sell it. This implies that the money can actually be used at both ends of the transfer chain. This permits multiplying the means of payment and increased liquidity. The form of the asset along the chain can be changed to anything: Cash, loan given, placement or others.

Note: Creating money as means-of-payment is not creating wealth. The money may well come from retained savings but equally can it be created where trust and confidence are abundant enough. The parts of that asset – liability pair have equal values. When all those loans are paid and the money disappears again it will not make anybody poor. When a liability gets obeyed and an asset is paid back they disappear at our institution. In the originating institution the assets returns to some other form, say from loan given to account balance or cash. The only income and only source of wealth here is the interest rate margin between the different loans. If there’s trouble with repayment but still confidence enough then it may be a good idea to throw good money after bad money and prolong the process with a new loan. But finally if some parties involved are not worth the trust but will default then some others will suffer and will lose actual wealth. The requirement is that a party who has received money must duly return them. A chain of defaults will destroy existing real property of participants. This may reach up to the asset at the beginning and beyond.

For the bank business all the previous is essentially technical arrangements. The income comes from interest rate margin and transaction fees. It’s good to attract depositors as they get a smaller rate than the interbank rate. Corporate borrowers may provide lucrative margins but the loans may turn soar during recessions. Mortgages for houses are good especially in a legislative environment like Finland where the position of banks is contractually strong. Any creation of endogenous money must be subject to those actual business requirements.

*Kydland & Prescott observed in 1990 empirically that the difference M2 – M1 leads financial slowdowns: “.. if anything, the monetary base lags the cycle slightly …The difference of M2–M1 leads the cycle by even more than M2, with the lead being about three quarters.” This matches well with the notion that money market issuances are needed to create long lasting endogenous money.*

The claims of post-Keynesians, that loans can be provided out of thin air by just extending bank balance sheet seems to be, while not exactly true, but extremely close to reality. A good review against various functional predictions is in UnlearningEconomics. +A major presentation of the concept of endogenous money was in Basil Moore’s 1988 book Horizontalists ans Verticalists.+

*A surprising feature is that this is not limited to home currency. If some Finnish banks want to create significant amounts of USD or CNY they can do that provided there are not legal limits on such activity.*

Effect on Macroeconomic Theory

Existence of endogenous money changes certain basic assumptions and rules. One is called money quantity theory. The idea is to treat money and interest rates as commodities. If more money is available then interest rates will go down and vice versa. That can’t hold. Most often we must look at central banks as rate setters. Further on major parts of monetarist theory. The assumption is that quantity of money creates economic expansion or inflation. Then we have the Keynesian branch established by Hicks. In IS-LM diagrams the interest rate is assumed to depend on money quantity which can’t be true. All those theories must either be abandoned or re-established on some new foundation. For IS-LM there exists that kind of approaches.

The post-crisis attempts in bank regulation are pathetic. Arbitrary amounts of share capital can be created in circular setups.

If we think of money as a stream flowing from central bank to the real economy then in fact it will not behave as a simple waterfall. Instead on the route there are great springs that occasionally provide large additional flows. During a recession it’s more like a swamp: The heavy deleveraging causes those former springs to operate as wells that drain the flow. It would be fine to see more empirical and theoretical description on such processes.

*Disappearing of money may possibly be a major reason for lack of demand in the initial phase of depressions as business financing will be cut back. This may cause basically good companies to go bankrupt. Another key factor to start a recession is that borrowers realize that the bubble created by lending is collapsing. Then failure to provide adequate stimulus leaves the economy in a depressed stable state with too little demand. The term liquidity preference is vague as actually its lending aversion. Ultimately there will appear new growth even after such economic mismanagement. In a modern world the recovery is, however, much slower than it was about thirty years ago. New companies must be created so that they are best in the world in respective areas. Such restrictions as the 60% state loan limit in the European stability pact may make it impossible to circumvent this kind of instability. State run banks might help to some extent.*

Copyright @ 2014 by Olli K. Ranta


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s