Bank income and spending – Part 2

John Hermann

It would be helpful to amplify various points on this topic made within the previous issue. Under the heading “Bank equity is not money” a distinction was made between bank financial assets and bank equity (neither of which are money, as used in the real economy). The financial assets of a bank are its reserves, investment securities and loan securities. As mentioned, the investment and loan securities are also liabilities of the borrowers and/or security issuers.

However in practice the aggregate of these bank assets is not matched by the aggregate of bank liabilities (bank equity is defined to be a measure of that mismatch).

I will give an example of why it is not possible, on a bank’s balance sheet, to match a specific asset with a specific liability, even if that liability was created in conjunction with the creation of an asset. Suppose that a bank (which may be called the lending bank) advances a loan to a retail borrower. In the overall process the lending bank creates a loan security as its new asset, and a deposit of credit money for the borrower as its new liability. However suppose that the borrower now spends some of that money into the economy, entailing transfer of part of the deposit to the custody of another bank. What happens is that the lending bank’s new loan security remains unchanged but part of the lending bank’s stock of reserves is commensurately transferred to the other bank. It is clear that one cannot talk about a matching of assets and liabilities. All one can say is that the aggregate of the lending bank’s assets has been reduced by the same amount as its reduced liability, so that its equity remains unchanged.

Comments from Jamie Walton

One of our readers, Jamie Walton, has commented on a few other points in the first article, As a matter of interest we reproduce his commentary below, along with my responses.

1. JW I’ve never heard of a bank’s “operating account” before – I’ve never seen it on a bank’s balance sheet in annual reports; is it only a notional account? [NB: Former banker John Tomlinson says that banks don’t destroy money when bank loans are repaid, they put it into their own account.

I don’t see that this could be the operating account you describe, but maybe it is another notional score- keeping account in which banks record by how much they have to increase their balance sheet again to replace the loan repaid in order to maintain their market share?]

JH For both banks and nonbank organisations an operating account records assets, liabilities and equity, as well as incoming revenues and out- going expenses. For many purposes it can be taken as synonymous with the term balance sheet. It differs from a transaction account, in that the entities recorded in it are not necessarily a form of money. Bank equity is not money for the reasons previously stated, while for a nonbank its equity can be (and often is) money. Banks do not undertake retail trading using their financial assets (reserves, loan securities, investment securities).

Tomlinson is correct in a very restricted sense. Banks have no need to actively destroy “money” when loans are repaid, for the simple reason that whenever a retail loan is repaid to a bank the money supply M1 is temporarily reduced – via the accounting conventions. Bank credit money is increased and reduced every minute of the day. It is true that reserves are transferred between banks when retail loans are repaid, however reserves never form any part of the money supply.

JW Yes, it seems that the “operating account” is the entire balance sheet, and when banks engaged in proprietary trading they are using the banks’ balance sheet to do it – presumably buying assets by either:

1. creating deposits to buy assets from their own non-bank customers,
2. transferring reserves to other banks to buy assets from other banks, or from other banks’ non-bank customers,
3. transferring deposits balances of the bank held with another bank to buy assets from other banks and non-bank customers with deposit accounts at that other bank, or
4. having some of their deposits at another bank destroyed to buy assets from that other bank.

(It can get a bit more complicated with international transactions, and there are about 4 ways these can be done, often with more intermediate steps, but as I understand it, the mechanics of the 4 options listed above are basically the same)

2. JW You wrote: ” a deposit (of credit money) in a bank is not a loan to the bank, as some people have been led to believe, because anything that is borrowed is necessarily an asset of the borrower.” This of course makes perfect sense. However, legally, well- established case law has deemed that bank deposits are loans of money by depositors to banks; generally unsecured loans at that, and thus bank depositors are unsecured creditors of banks.

JH I am aware of such legal rulings. The law is an ass at times, and many magistrates (amongst other sectors of society) are simply ignorant of both the accounting reality and the underpinning logic. I would point out that deeming something to be what clearly it is not does not change the reality.

3. JW This was based on a scenario where a customer deposited physical cash with a bank. Since banks promise to pay out cash at any time on any demand deposits, it seems that specific scenario has been extended to be the general scenario, and all depositors are deemed to have deposited cash with a bank, regardless of how those deposits were originated. [This seems to also be the mainstream view of economists also, e.g., the much-vaunted Diamond- Dybvig model, which assumes that the (single) bank is waiting for investors to invest illiquid assets with the bank in exchange for liquid demand deposit “contracts” before the bank can in some way make illiquid loans of (something of) long-term maturity (This model has become the basis of mainstream banking theory ever since it was first published in 1983)]

JH When a person walks into a bank building with currency (coins and notes) and asks for a deposit to be made in his/her account, a magical transformation occurs. Upon receipt the currency ceases to be part of the money supply M1 (the aggregate of money held by nonbanks) and is transformed instantly into being part of the bank’s currency reserves (which are basically inter- changeable with its creditary reserves). The aggregate of the bank’s reserves temporarily increases and the money supply remains unchanged because a creditary deposit (of bank credit money) has also been created in the customer’s account to counterbalance the loss of currency from M1. This is a transformation, but it should not be thought of as a mixing of reserves and the money supply. That line is never crossed. The reverse process occurs whenever a customer makes a withdrawal and demands to be paid with currency.

JW I agree. The point I was trying to make is that the same physical cash can go from a central bank vault (or Treasury vault, in the case of Treasury coins and Treasury notes) to a bank vault to a bank customers’ wallet, and back again (proposals for government- issued digital tokens, e.g., in China, can enable the same thing to be done digitally, possibly all on one ledger).

Also, in the U.S. there are about 250 non-banks that have accounts at the central bank, including financial market utilities such as the DTC (not that long ago, central bank staff had accounts with central banks). There is a movement in Europe to allow anyone to have accounts at the central bank, e.g., in the Netherlands (where they got a unanimous vote for it in Parliament), and many/ most big companies, e.g. Siemens, have their own banks so they can have accounts at the central bank (although usually unstated, my understanding from investigation is that it’s because they don’t trust the solvency of the big commercial banks, and want to avoid a “bail-in” of their funds by having their funds held at central banks (which, presumable, won’t have to do a “bail-in” of reserves)). (Hopefully, this will all tend to lead to the system being reformed sooner rather than later.)

JW You wrote: “Reserves held by commercial banks are not part of the money supply, and neither are they interchangeable with bank credit money”. However, “cash held in bank vaults and tills (or currency reserves)” is interchangeable with bank credit money (and central bank credit money).

JH The statement that I made immediately above can be summed up in the following:

(a). Bank credit money is interchangeable with currency held by nonbanks. (b). Currency held by banks is inter- changeable with reserves.

(c). Bank reserves (including bank currency reserves) are not interchangeable with the money supply.

If (c) were not the case, then nonbanks would have access to reserve deposits (i.e. they could have central bank accounts) and banks would be able to lend out their reserves. But this is currently not the case. That is, banks at this point in time do not lend out their reserves to their retail customers, they only lend out newly created bank credit money. The reserves tag along with bank credit money with every transact- ion involving banks, without mixing.