Friday, 7 October 2011

What proportion of the money supply is created by commercial banks?

I like the ideas in this paper authored by Prof.R.A.Werner, Positive Money and the New Economics Foundation. But I’ve got a technical quibble about the proportion of the money supply that is created by commercial banks.

Pos Money and NEF argue that because about 3% of the money supply is physical cash (£20 notes, etc), therefor about 97% must be commercial bank created money.

That point would be valid if physical cash was the only form of monetary base. But it’s not: there is also book keeping entry type monetary base: i.e. bank reserves.

NEF claim that “central bank reserves do not actually circulate in the economy” which is their reason for not counting reserves as part of the money supply. I beg to differ, and for the following reasons.

Suppose I have government debt which reaches maturity tomorrow. I’ll get a cheque from the central bank tomorrow for the relevant amount, which I then deposit at my commercial bank. And the latter than presents the cheque to the central bank, which then credits the account of the commercial bank in the books of the central bank.

I can then transfer £Y of my recently boosted bank balance to person X. And the central bank, as part of it’s role in helping commercial banks settle up between themselves, will transfer £Y from my commercial bank’s account at the central bank to the account (in the central bank’s books) of the commercial bank where X keeps his money.

Now the above transfer is no different to the transfer of commercial bank created money between two individuals (and hence between their commercial banks). Put another way, when I get monetary base as a result of my government debt maturing, I then in effect have money stored at the central bank. As to transferring this money to someone else, it just happens (because of institutional arrangements) that commercial banks act as “go betweens” when it comes to transferring this money. But effectively, the chunk of monetary base concerned is part of the money supply “circulating in the economy”.

But that doesn’t detract from any of the basic points made by Pos Money or NEF. It’s just that I think their 97% figure is a bit out. Since bank reserves are roughly 3% of total money supply (far as I can see – which is not very far) I am saying the figure should be about 94%. Though at the moment, bank reserves are somewhat bloated, so the figure will currently be less than 94%



  1. There's a very strong argument that Gilts should just be seen as part of the 'money supply' since they are really just disguised Bank Reserves.

    And then there's an argument that you can add any deposit guaranteed by the state - on the fairly reasonable assumption that it wouldn't be there other than for that guarantee but in National Savings instead.

    The 97% is truth by repeated assertion and frankly is a silly argument.

    The debate is one around who should decide how much money is required by the system. Is that a centralised decision or a private decision.

  2. Ralph,

    I believe that both, you and NEF are wrong on that. Accumulated Government deficits, or so called National Debt is Government created money, and not created by commercial banks. So from M3, I would subtract the National Debt, and then take the ratio of that to M3. Please correct me if you think I am mistaken.

    If I look at two time series M3 and FGSDODNS (Federal Govt Debt Outstanding), and take the ratio, then in 1980, government money was about 35% It increased to about 80% by 1996, and then declined to 40%+ by 2002, and then increased to about 45% when they stopped publishing M3. The graph can be seen at

    My thinking goes as follows

    1) Government creates money when it spends
    2) Government destroys money when it taxes
    3) Therefore accumulated deficits indicate Government created money in circulation or saved.
    4) Private banks create money by issuing debt.
    5) Private banks destroy money when debt is repaid or written off.
    6) Accumulated private sector debt is money saved or in circulation.

    Sum of these should be M3 or something close to it. It is possible that the real money supply is a bit larger than that.

  3. Hi Clonal,

    Using conventional definitions of “money” and “national debt” it is not true to say that “National Debt is Government created money…”. I.e. when the government/Treasury net spends, it can only do that by borrowing. Thus it is national debt that rises when government net spends, not the money supply. The central bank may then print money and buy some of that debt, or it may not.

    However, there is no sharp dividing line between money and non-money. Thus as both Neil and I suggest above, national debt could at a stretch be counted as money. However, for the purposes of my post above, I was using conventional definitions. So using conventional definitions, your M3 / FGSDODNS ratio is not really relevant to the point I was trying to make in the above post.

    The Fed’s M3 and monetary base figures actually support my point – sort of. That is, the fraction “monetary base on M3” hovers roughly around rougly 10% between 1980 and 2000. See:


    I actually said above that bank reserves (i.e. monetary base) were normally 3% of total money supply in the UK. I’ve no idea why there is such a discrepancy between the 3% UK figure and the US 10% figures. I can’t remember where I saw the 3% figure: it might have been an unreliable source. But if 10% is the figure for the UK, then that is all the more reason for my difference of opinion with the NEF to be sorted out.


Post a comment.