The Equity Residue in Bank Deposits

A short note in reply to @Frances_Coppola. She disagreed with a paper from Jan Kregal at Levy.

Kregal argues that there are two types of deposit, deposits of currency and coin, and deposits created when loans are made.  If a bank makes bad loans

“it is the failure of the holder of the second type of deposit [loan-created deposits] to redeem its liability that is the major cause of bank failure”

so the first type of depositor (of currency and coin) should not bear the brunt of these bad decisions.

Coppola disagrees with this on a theoretical point via twitter.

deposits physically deposited by customers are receipts of deposits created via someone else’s borrowing.

Here we agree to a point, reserves created by loan are spent in excess of the liquidity preference of the borrower and such ‘excess reserves’ creates reserves in other bank account and further expands the lending power of banks to lend on these reserves. We have modelled in detail this process before, as did many other banking theorists when the endogenous ex nihilo creation of money was taken for granted in banking textbooks.

But only to a point because there can be two sources of deposits (setting aside for a moment any debt free state created money) as Kregal correctly points out – Crusoe like savings from balances in excess of liquidity preference – and the endogenous creation of moneys through loans.

This must be true as balances immediately spent are not available as idle balances to be leveraged as lending. Also imagine a bank starting up, it cannot start up without equity, similarly a bank cannot expand beyond its capital ratios without raising additional equity.

This arises from the fundamental equation of accounting applied to the business model of banking, Assets=liabilities+owners equity. When a new bank, or a bank extending its capital base, creates a new loan it creates a liability now and an asset redeemable only in the future, to ensure positive balances the bank must raise equity, or its own liability through borrowing from another body, the two sides of the equation must balance. If it borrows the bank reduces its profits, hence why banks borrow short. This in no way implies a loanable funds view of money, rather future profits have two sources, saved existing money and loans, invested. It is the anticipation of the future stock of such money, not the present stock, which constrains lending decisions.

So even with the ex nihilo creation of deposits for loans, which create deposits for loans etc. etc. their is always a residue both of the original equity and the created money. In banks with high reserve requirements it will be higher but there will even be a residue in regimes with no reserve requirements.

This argument is formally the same to that of Bose (1980) in discussing the classical concept of ‘originary factors’ – and the ‘dated reduction’ of originary factors. Every commodity is composed of another commodity (a resource) and an input of labour. Going back through time we find natural resources and labour. This means that no commodity can be ever reduced to ‘pure labour’ or ‘pure commodities’ there is always a residue of one or the other. (Mathematically it is a Taylor series where ratios between the two are invariant over time). Here we have not referred to ‘capital’ rather purely the physical economy. However the circuit for the physical economy is paralleled by the reverse circuit of money. For the classicals ‘capital’ was simply money advanced for profit, including money advanced for wages and interest (quite correctly), and this money also has two sources, savings and loans. This leads to a rather interesting flow identity for a specific investment, prices, at effectual demand,  equal:

commodities + labour = change in saving + change in debt

It does not matter if the change in saving is from retained profits (self funding) or change in savings to create bank equity and further capital for lending.

Update:  I should have stressed here the LHS is the full cost price of production and the RHS is what the later classical economists (such as Tausseg) called the ‘pool of funding’ both measured as flow variables over (Wicksell’s term modifying Bohm -Bawerck) the ‘Period of investment’.  Also we are talking about the change in savings and debt that is spent not retained in idle balences, so both are multiplied by the Davidson k factor (the proensity to hold income in balances).

But the value of labour here is simply the labour share (α ) times velocity ( α .v )which is the same as (1-r).v (r=the profits share, v = the velocity of money).

So we have a flow equation

Commodity price =(Δequity+ Δdebt)/ (α .v)

You can of course rearrange this equation to make endogenously created bank debt the dependent variable, but you cannot ever eliminate saving on the RHS, there is always a residue. These equations should come as no surprise, they can be derived from Kalecki’s stock identities modeled in a flow input, flow output system.

Two other interesting things about this identity. Declining labour share is inherently deflationary, indeed during the great moderation we saw deflation and declining labour share, and incomes and profits can only remain even with increasing debt, increasing savings for equity is also deflationary.

So what happens when labour share falls towards zero in an ‘android’ economy. I will explore this in a future post (it isn’t pretty).

(Note: At no point do I assume the such dated quantities can be ‘added up’ to equate values, Sraffa demonstrates the problems in this. Rather I assume that all such inputs are valued at current prices).

Further Reading


Marx on Exploitation and Inequality: An Essay in Marxian Analytical Economics


Bose, Arun

The Transmission of Lending Power between Banks in Endogenous Monetary Theory Lainton

Correctly Modelling Reserves, Cost of Funding and Collateral in Monetary Circuit Theory Lainton

Production of Commodities by Means of Commodities Sraffa

The Economics of Enterprise (HJ Davenport) 1813

Collected Works, Kalecki