Business and Society Economics Money and Banking

A Banking Debate

Receive Email Updates on New Posts

Wall Street, NYC
Wall Street


Since the financial crisis of 2007-8, one suggested target reform has been the monetary system itself.  This reform is based on the recognition that money in the modern economy is a rather peculiar phenomenon.

There are two popular conceptions of the nature of money, both of them incorrect.  (Note that when we talk about money, it is entirely artificial to separate cash, in the form of bank notes and coin, from what we hold in bank accounts. To all effects and purposes, for the vast majority of us, they are the same and completely interchangeable.)

The first conception is that money is a fixed quantity determined by the government, which is either accepted by convention or because you can go to your bank and get a certain quantity of gold for it. (Presumably not many people have actually tried this!) The second is that banks can issue new money to lenders as a multiple of pre-existing deposits, depending on how often depositors demand cash. This is frequently referred to as ‘fractional reserve banking’.

Assets and Liabilities

The reality is that while commercial banks can indeed create money, reserves are only an indirect constraint rather than a direct one. This is because when banks issue loans they do not need any deposits to do so – because the act of creating a loan also creates deposits of the same value. This only makes sense once we consider money as a financial asset which like all such assets is also a liability. Financial assets are both asset and liability because they are IOUs. They are a liability to I and an asset to U. This explains how money can have value although the paper and ink or electrons which embody it are almost worthless. The paper or computer entry is a record of an obligation from one agent to another – from the issuer to whoever currently has the title of that record.

When a bank issues money it is an asset to whoever it is lent to, and a liability to the bank.  Why would anyone willingly issue a liability on themselves? The answer is that it permits the bank to create an asset of its own. This asset is the loan which itself represents an obligation to repay (thus a liability) for the loan recipient – the borrower.  Since all the new liabilities and assets net out, no new financial wealth has been created.

Why is there any motivation for lenders and borrowers to do something which in accounting terms as zero consequences? While the bank believes it will make a monetary profit because of the interest flows from the loan, the borrower believes that by bringing forward the ability to make purchases he or she gains a real production and welfare advantage.

Limiting the Creation of Money

The failure to understand money’s role in terms of balance sheets where the credit and debit aspect of financial assets are clearly laid out leads to much confusion.  I believe this was at the heart of a recent debate I had with a monetary reform campaigner. The nature of the proposed reform he supports is that commercial banks will no longer be allowed to create money when they lend. This makes the first misconception about money become true! The argument for this is based on the correct observation that increasing bank lending tends to accompany ‘booms’ that then lead to ‘busts’ exactly as was observed in the 2007 financial crash. Take away the bank’s ability to lend autonomously – the problem goes away (so the argument goes).

There are three main objections to this reading. Firstly, correlation is not causation – so it may not be that the ability of banks to issue new money and loans is the cause of booms and busts although it is clearly part of the pattern.  If it’s not the cause, the same problem may still occur, but just in a slightly different guise. Secondly the reform may create as many or even more problems than it solves.  Thirdly, in a way an aspect of both of the other two objections, is that the creation of exchangeable IOUs is not something that can in practice be legislated out of existence, but only driven into the shadows.

Most of these points (particularly that limiting money creation would be contractionary and restrictive) were made in a post by Ann Pettifor that was the starting point of this debate. These were summarily dismissed by the proponent of reform, on the basis that the restrictive effects could be offset by tax cuts, that government money issue would avoid people being ‘chained to debt’ and that Pettifor didn’t like the proposal because Milton Friedman had proposed something similar in the 1920s and 30s. Positive points made were that under the reforms banks would become financial utilities rather than creators of too much money in booms and too little in busts, that it would be possible for the government to specify what money is and how it is measured, as well as how much is needed and then distributed through government spending, tax cuts and rebates, and distribution to local government etc. It was also argued that this ‘debt-free’ money could be used for the removal of the national debt and to avoid the interest repayments (sic) that consume the tax base.

Private Plans and Government Debt

I responded at this point with the arguments that the restriction of decentralised money creation could prohibit otherwise productive plans agreed by lender and borrower, by making lending less responsive and targeted to where these plans existed, and that the issue of money by government put the burden of risk of issue on all citizens rather than primarily on the borrower and lender. The response: the government can always lend money to banks and people will ultimately lend money they have saved. Moreover, decentralised lending cannot be controlled by a natural interest rate set by the central bank.

I argued in response that the issue of money to pay off government debt doesn’t really achieve much for a nation with its own currency – it simply exchanges one liability of government for another. As it is, quantitative easing (QE) represents exactly the process described, except that it is targeted at banks themselves (it could be otherwise) and the bonds purchased are not eliminated but remain on the Bank of England’s balance sheet as assets to the Bank but liabilities to the government. In response to the problem of interest I argued (as here) that the interest payments from borrowers to banks are simply a repeated circular flow: from borrowers to banks as interest (allowing banks to pay employees, purchase capital and pay shareholders’ dividends); from the banks’ payees back to borrowers; and from borrowers to banks again, this time as repayment of loans.

The government may be able to lend to banks, but this dilutes the effect of the reform while introducing new problems. The government (or its dedicated agency) must now decide whether such loans are justified while taking on the risk associated with those loans. Even if the government creates as much money as commercial banks would have done (which seems to go against the argument for the reform), for the money to find its way to all bilaterally agreed (between lender and borrower) projects assumes that those not spending are also willing for their savings to be lent out on those specific projects – but there can be no guarantee of this.

Government Money as Citizens’ Risk

What is the nature of the risk of government issued money? Firstly if issued as loans to commercial banks, then there is the risk, as with all loans, that they may not be repaid. If they are not, this represents money intended for temporary circulation that ends up in permanent circulation – thus there is a risk of inflation that devalues all deposits and fixed-income assets. If the government adopts the same accounting method as commercial banks currently use, which is to write off bad loans from their equity then this reduces the government’s future spending power – at the expense of all citizens. For money issued directly – as spending or other payouts, the risk is that too much will be inflationary – requiring increased taxes, additional debt sales or resulting in the erosion of savings again.

The debate now turned to the nature of the money the government currently spends. It is my contention that when government spends money it creates this money, subsequently reabsorbing it to avoid inflation through taxes and debt sales. This requirement on citizens to pay taxes is also part of what gives money acceptability and determinate value.

The counterclaim was that money spent by the government has been previously created by commercial banks and acquired by government by taxes and bond sales. The importance of this claim is that it would mean putting us in hock to bankers so as to allow us to pay our debts, and government in hock to the beneficiaries of money creation. Of course the flow of money appears to be perfectly compatible with either claim – since money flows from banks to citizens and from citizens to the government and then from the government to citizens and back to banks in either case. To identify what is actually going on we need to consider the logic of balance sheets. The money exchanges between the government, the central bank and the private sector take place entirely through the medium of reserves, or ‘high powered money’. This money can neither be created by banks, nor used to make purchases in the real economy. But when it is transferred from the government to the asset side of a commercial bank’s balance sheet it allows the commercial bank to costlessly create deposits in the name of the government’s intended payee. In fact not only is it costless for the commercial banks to accept such payments, it is an advantage in that it has the effect of increasing the commercial bank’s reserve ratio (as long as it was not 100% to begin with). Any subsequent transfers of deposits to other commercial banks are accompanied by a transfer of the same quantity of reserves – thus maintaining the balance sheet status of donor and recipient banks. When taxes are paid or government or purchases made, the equivalent quantity of deposits and reserves are eliminated from the commercial bank of the payer – again maintaining balance sheet status. Note that while none of these transactions affect a bank’s balance sheet status, they do have consequences for reserve ratios – and since the transactions themselves require reserves they may lead to some banks having a shortage and others having a surplus of reserves. This can be rectified in two ways – firstly surplus banks can lend to shortage banks; secondly shortage banks can borrow at interest and against collateral (usually government bonds) reserves from the central bank. If the banking sector as a whole is increasing its lending relative to loan repayments (and thus the quantity of money is increasing) then there will generally be a continual demand for additional reserves from the central bank. This may give the central bank some influence on the rate of commercial bank money creation. It almost always tries to implement this control by being willing to lend an unlimited quantity of reserves at a fixed price (the bank interest rate) rather than imposing a limit on the quantity of reserves lent.

Balance Sheet Analysis

What is the evidence that the government can spend independently of tax and debts rather than the alternative where it taxes and borrows commercial bank money and subsequently spends this back into the economy? The first exercise is to look at bank balance sheets. We see deposits on the liability side of the commercial banks’ balance sheets and reserves on the asset side. So if the government makes a payment into the bank account of a state employee or the recipient of a state pension it must be adding to bank liabilities. Is it plausible that this occurs without some accompanying movement in commercial bank assets? For if it does not the act of the commercial banks in accepting such a payment into the account of a depositor reduces the net worth of that bank – indeed a large enough quantity of such payments could lead to the insolvency of the bank, where its liabilities exceeded its assets! Conversely a payment to government from a depositor, in payment of tax or for the purchase of government bonds would reduce liabilities of the depositor’s bank with no changes in assets – the bank’s net worth would increase! One option is that banks could make up this shortfall by borrowing reserves from each other and the central bank. But there would remain positive costs in accepting government payments and benefits to making them. Banks would surely react by asymmetrically discouraging depositors who received government payments and encouraging those who made them. I am not aware of any evidence that banks attempt to discriminate in this way.

A similar argument was made against reserves being a balancing element in transfers between commercial bank deposits. The argument was that rather than there being an accompanying shift in reserves with any transfer of deposits – when deposits are transferred the recipient bank makes a loan back to the ‘donor’ bank.

While this keeps both banks in balance it has the effect of reducing the second bank’s reserve ratio. While it is difficult to prove that this does not happen, it must be less plausible than the simpler idea that reserves are transferred with deposits.

Keynes’s point about the ability of banks to create money if they do so ‘in step’ is quoted. But ‘in step’ implies that they do so at the same rate and that new deposits are distributed evenly – otherwise there is the risk that net transfers or cash withdrawals will exceed available reserves in aggregate and the banking system becomes illiquid.


The existing system, while it does allow banks to create money purely by accounting entries, they do so by creating both an asset (the loan) and the liability (the obligation initially to make payments on behalf of the depositor and ultimately to enforce repayment of the loan or suffer a commensurate loss through their profit/loss account). This loan and money issue is a bilateral agreement with the primary risk borne by the borrower and the lending bank. The direct risk is only transferred beyond this bilateral relationship if the bank is unable to meet the equity loss from failed loans and is deemed too big to fail (TBTF) by the overseeing government. The losses that eventually result from such ‘bail-outs’ are usually much smaller than the initial guarantees that have to be made to keep the bank trading.

Currently the government does issue money for its own expenditure and it taxes and issues debt not to acquire money to spend but to maintain the quantity of reserves in circulation that allows the desired measure of control of the price of reserves. The balance between taxation and debt issue for this purpose is not driven by any shortage of money, but by incentive/disincentive concerns over taxation, smoothing of tax rates, the requirement to provide safe assets for financial institutions and political imperatives toward special interests.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.