‘Forced Savings’, the Credit Accelerator and the ‘Revolving fund of Finance’
From the models we have produced in the previous weeks we are finally in a position to begin to tackle some of the fundamental puzzles and controversies of monetary theory. Some of these controversies being forgotten in the neoclassical literature because of assumptions of money neutrality, equilibrium and the unimportance of credit/debt.
From our quadruple entry models of banking and investment we have shown that even within a framework where by definition (in the Keynesian sense) investment creates savings there are noticeable lags and periods over the term of a loan where they are not equal in terms of invested funding=additional balances, they balance only over the term of a good loan. We have also shown that turnover matters and because of this excess reserves matters as does the turnover period of a loan. In a sense we have extended the inquiry that Torrens began, that if turnover matters and profits equalise through capital markets then what must this mean for prices? We have extended this approach to banking and what it means for growth. The most striking result concerns our investigation of Keyne’s concept of a ‘Revolving Fund of Finance’ we have demonstrated that this can grow and service additional loans without any prior savings (in the sense of funding) once it begins profitable investment. But in a striking result we have shown that this fund can only commence and expand above this ‘natural rate’ through prior ‘crusoe’ type savings. A result that cuts rights across the debates of Keynes and Hayek – in a sense they were both partially right on this issue.
I want to explore this further. One of Hayeks key obsessions was over ‘forced savings’ it is central to his concept of the business cycle and capital theory. It is founded on the view that credit fuelled growth matters to pricing structure – it is a particular case of the Cnatillon effect. This is intriguing because the rediscovery (I word I use deliberately) by Keen, Mayer et .al of a Credit Accelerator – a concept that undermines Says Law – is on exactly the same basis.
Of course the mainstream view has been that credit doesn’t really matter for pricing and business cycles (if it doesn’t matter for pricing then by definition it cannot matter for business cycles). The point I think was most strongly put by Ricardo in 1819 during parliamentary cross-examination on the bullionist controversy.
Credit, I think, is the means which is alternately transferred from one to another, to make use of capital actually existing; it does not create capital; it determines only by whom that capital should be employed…Capital can only be acquired by saving 
Hayek certainly held that capital could only be created by prior ‘Crusoe’ type saving, but he disagreed on the issue of pricing extending the classical idea of ‘forced savings’.
The idea of forced savings has been poorly framed. Machlup detected no fewer than 26 different uses of it. Schumpteter thought the term unfortunate and confusing.
The form from which Hayek obtained it comes from late C19 economist Thomas Joplin – a key figure (founder of the Currency school and to my mind the Austrian school as well) in that it is from his conception of ‘forced savings’ that Wicksell adopted the concept of a ‘natural’ rate of interest. (Viners treatment in Studies in the Theory of International Trade is recommended). Joplin:
If a person borrows one thousand pounds of a banker who issues his own notes, the banker…has at once added a thousand pounds to the capital and a thousand pounds to the currency of the country. To the party who has borrowed the money, he has given the power of going into the market and purchasing a thousand pounds’ worth of commodities, but in doing this he raises their price and diminishes the value of the money in previous circulation to the extent of one thousand pounds, so that he acquires the commodities by depriving those of them who held the money by which they were represented and to whom they properly belonged. On the other hand, if a person pays a thousand pounds into the hands of a banker, and the currency is contracted to that extent, both one thousand pounds of capital and one thousand pounds of currency are destroyed. The commodities represented by the money thus saved and cancelled, are thrown on the market, prices are reduced, and the power of consuming them is obtained by the holders of the money left in circulation
You will note here a completely ‘horizontalist’ position. The expansion of the monetary stock has a 1:1 impact on prices. Joplin did not consider this a good thing.
Legitimately a banker can never lend money which has not been saved out of income. Money saved represents commodities which might have been consumed by the party who saves it. Interest is paid for the use of the commodities and not for the money.
And Viner comments
If banks have the power to issue money, the amount of such issue is determined by the rate of interest which the banks charge on loans. If forced saving is to be avoided, banks should charge “the natural rate of interest,” which he defines as the rate which keeps savings and borrowings equal.
For Joplin the quantity of money,
“which ought, if possible, to be as fixed as the sun-dial, came to depend upon the credit of bankers with the public, and the credit of the public with the bankers.. which ought no more to affect the amount of currency in circulation than the motions of the sun.”
This of course is exactly Hayek’s and Mise’s position, and setting aside the complications of their theory’s on capital structure and the business cycle this is the analytical core of their system. Bank credit alters the price structure. Hayek added to this what he called the ‘Ricardo effect’ that is with relative labour costs increased (because of the increase in prices of consumer goods) there would be capital/labour substitution which would alter the capital structure.
In an echo of the closely related classical concept it creates ‘fictitious capital’ or in his terminology ‘malinvestment’. Because of the dogma that only savings create capital it is termed ‘forced savings’ when it in fact is not savings at all but disaving, being forced to run down idle reserves because of a price increase. Note this is the exact opposite of the opposite dogma of Hahn (and Rae and – in part- Schumpter) that only credit can create capital.
[Note there is another partially related use of the term 'forced savings' in the literature relating to behaviour when credit is created under conditions of full employment. This is the only use of the term treated by Keynes. We are not concerned for the time being with this usage here but the broader concept used by Joplin and Hayek which applies in their schema throughout the entire upswing of the business cycle].
Hayek builds on this but does not alter this essential foundation. For Joplin the objection was distributional, wage costs would rise before capital saw increased profits, debtors would receive income before creditors would receive interest. There is little of this in Hayek or Mises for whom the objection relates to the alteration of capital structure.
The first point to note about this theory is that it is wholly incompatible with a ‘pure’ time preference theory of interest. Although the Austrian Theory of the Business Cycle is often presented as one of where Central Bank induced distortions above and below the ‘natural rate’ drive the cycle, but in reality, as is very clear in its original form from Joplin, the objection is to fractional reserve banking, the use of credit and the expansion of credit (lending power in our terminology- which excess reserves allow) – which is the issue. This is the presentation in the more thoughtful Austrian treatments (such as at the Mises Wiki).
If it is the case then that the interest rate is influenced by monetary ‘distorting’ factors then it is not set by pure time preference, there is an additional influence.
On this issue Joplin, was very clear, clearer than Mises (in the Theory of Money and Credit) or Hayek (In Prices and Production, Hayek later modified much of his monetary approach), in his view the interest rate was not set by the supply and demand for saving but by the supply and demand for money. It was a theory well ahead of its time – in essence Keynes Theory of liquidity preference even including a breakdown of the motivations to hold money at any one point in time. So for Joplin it was an imbalance in the demand for money, caused by an imbalance between savings and investment (though by no means the only potential imbalance) which caused the interest rate to vary above or below its ‘natural’ rate which he defined as that being as the rate design to keep savings and investments in balance.
In Hayek we can see a strong assumption that all monetary distortions that change price structures away from the relative values of a pure barter economy are a bad thing. The assumption is that it is the excess demand for money (in Walrasian terms) which creates disequilibrium and drives the business cycle. Whatever the weaknesses in Hayek’s approach due to his anti-credit bias this is an important insight.
However if all investment is forced to come from past capital accumulation then that presumes a past and steady period of capital accumulation to fuel future economic expansion. This deeply ‘English’ assumption was challenged by many C19 political economists in America and Germany. Without a period of past capital accumulation how were they to compete? This led to schools of writing where credit and infant industry protection were viewed favourably. For example by the end of the century in Taussig’s writings we see the argument that when credit is advanced it is done so in anticipation of future profits. The key here being that a high profit rate, justifying a high interest rate, can compensate for lack of past capital accumulation.
This all rather begs a question – so does ‘investment’ need to be balanced by ‘savings’ in terms of express withdrawal of balances which could be used for consumption (which we more accurately term funding rather than saving)?
Joplin, in terminology later adopted by Austrians, saw ‘saving’ as simply being deferred consumption. From the perspective of the law of large numbers if ‘saving’ and borrowing are occurring at the same rate then the inflationary expansion of spending at the beginning of a loan period would be balanced by the deflationary contraction in spending caused by deferred consumption (as John Rae rightly saw it) towards the end of the loan period. This is also the way Hawtry verbally described the credit accelerator - in that a change in the rate of loans granted is necessary to have a net impact on effective demand and prices. The same approach can also be used to describe the impact on effective demand from investment funded from Crusoe type savings. At first saving is deflationary, but this as Austrians such as Strigl describe, simply builds up a ‘pool of funding’ for consumer goods during the period of production prior to sales. Again deferred consumption of consumer goods. Again the law of large numbers suggests that in a period of unchanging lending then the deflationary downswings are cancelled from the inflationary upswings.
The issue of inflation from credit has often been presented in terms of the special case of full employment where all labour is fully employed and where is ‘investment’ is fully funded by savings. (this is the second application of the term’forced savings’ in our note above). In that special case (described by Bentham and Machlup amongst others) any increase in credit must lead to a direct increase in demand for productive goods which with all factors fully employed must lead to a rise in prices, a rise in profits and a forced reduction in money balances from consumers.
In the more general case though there is no assumption that any factor is fully employed, merely that an increase in demand for a good may or may not lead to an increase in the price of the good depending on the elasticity of its supply curve. The key though is whether of not NET there is an increase of decrease in demand in terms of the credit accelerator. We have seen that if the rate of lending is static then the law of large numbers ensures that inflationary and deflationary pressures exactly cancel out. If the credit accelerator rises there will be inflationary pressure, falls deflationary pressure.
You might imagine then that this would be equivalent to a world where ‘savings’ and investments are in equilibrium – you would be wrong. It is at this point we can apply the results of our modelling. What this shows is that even if such funding exactly balances investment it does not result in in a neutral position in terms of prices, that is because the revolving fund of finance enables the exponential growth of lending power due to profits from interest and not returned as dividends expanding lending power. In that case with the supply of endogenous money increasing interest rates must be pushed down, even if ‘savings’ (funding) and investment are in balance. We can see that this must be so from our finding that changes in lending power = changes in savings for funding, so if lending power is increasing from retention of the ‘bankers surplus’ it is possible for interest rates to remain static if there is compensating dis-saving - a negative and balancing rate of change in saving.
Ill present a simple model. In the first we have a ‘frontier’ bank with limited initial equity $10,000 – lending in conditions of a high rate of profit. Lets say it fractionally levers that to $90,000 of lending power leaving $1,000 in reserves. Lets assume an interest rate of 7% of which the bank makes 5% profit. (for simplicity for the moment we are leaving aside inflation), let us also assume that the bank pays a 5% dividend recycling 95% of the banking surplus to lending power. I also assume that the k factor – that is the proportion of the new deposits retained in the bank but not spent is 0.05.
This produces the following:
|Lending Power||Interest||Profits||Dividends||Reserves||Excess Reserves||Excess Reserves Levered|
You can see from this can lending power overall increases much more rapidly than savings (in the Keynesian sense of unspent balances) because of the increase to the revolving fund of finance. Two issues to make the model more realistic. Firstly because of a wealth effect savings are likely to rise with income. Secondly their is a second order effect with excess reserves being placed in other banks – this is a single bank model – the extent of this will depend on the turnover rate of balances.
The effect of this is that banks, at times of steady growth, will – over time – have less and less need to attract ‘savings’ (funding) to fund loans – so they can afford to lower deposit rates and hence increase profits because of the increased spread between savings and deposit rates. There cannot be a stable period where savings (funding)=investment under endogenous money as because of the changing size of the revolving fund due to compound interest it is forever shifting. Also remember investment = funding x turnover – and turnover is affected both by the turnover period of capital and the amount of excess reserves. The relationship between savings and investment is a profoundly disequilibrium one.
But what if there were no revolving fund of finance – what say if the state taxed all bank profits at 100% and returned it to citizens as a citizens dividend. In the model the effect would be identical in terms of expansion of lending power as what would be dividends in bankers accounts becomes reserves in citizens accounts and the excess reserves can equally be levered through fractional reserve lending. This again is because of the crucial role of excess reserves, here all unspent balances become potentially available for ‘funding’ through fractional reserve lending. If an investment is made then the funding for that loan creates savings mostly in other banks that are available for funding. Whether that loan is made or the funding is successfully secured is another matter. The critical difference between this scenario and that of the previous paragraph is that without the revolving fund there is no systematic downward pressure on interest rates. Let us assume for the sake of argument that all loan funding comes from active asset purchases by depositors. Here we have a 1:1 relationship between the requirements to fund investment and saving. Yet even here this supports lending at 1/the reserve ratio – endogenous monetary expansion. A faster rate of money creation than the rate at which it is saved.
But what if the bankers surplus is not returned as dividends, well it doesn’t matter because it becomes excess reserves in another bank account and the impact mathematically on lending power is identical, it expands to the extent of the reciprocal of the banks reserve ratio.It makes no sense then to complain of ‘forced savings’ in terms of an expansion of money in advance of saving – there is no such thing. All increases in the rate of lending power must be preceded by an equal increase in saving (funding) – either directly through asset purchases or indirectly through excess reserves or banks own savings through retention of the bankers surplus (indeed this expansion of the term ‘savings’ – in the keynsian sense enables us to see the revolving fund as an example of prior savings as well). The ‘flow’ of money liabilities through loans must be exactly balanced by the inflow of assets through principal and loan repayments and savings to fund these deposit draw-downs, even in those cases where through fractional reserve lending credit is created many times in excess of reserves.
So a key finding here is that the normal operation of the credit cycle, during a period of growth, has a natural tendency to push down the interest rate due to the increase in lending power. This is exacerbated by the accelerator effects of investment, as emphasised by Sraffa, wealth from investment further pushes down interest rates lowering costs and increasing costs raising wealth etc. If the interest rate is falling then during periods of profit there will be an expansion of credit as the cost of servicing the credit is low.
With a tendency for the interest rate to fall even a constant level of saving will not lead to a constant level of prices – a conclusion also reached by Hayek on the basis that a constant level of savings would lead to an increase in output which would be deflationary with constant nominal spending.
Hayek and Mises set out there theory as a defence of the classical theory of growth – that capital is stored labour and land transferred for products, therefore if you wished to produce more in the future there had to be storage of unused labour and land (savings). However our theory shows that the classical theory can be defended without any notion of ‘forced savings’ because credit acceleration depends on savings, either direct investment or idle excess reserves, and even with fractional reserve lending with no credit acceleration cash inflows to banks from productive loans exactly balance cash outflows, there is zero net inflationary effect.
The reflexive bias against credit and fractional reserve lending then is without basis. It does not create per se inflation.
It is another argument altogether about whether or not it alters capital structure (the Ricardo effect) – although as it is investment per se which enlarges the pool of funding which provides the real wage and demand for consumer goods – whether crusoe type savings or credit – this argument must fall as well. Such complication capital theory nonsense is not necessary to sustain the classical theory of growth – simple stock flow analysis of cash flows is sufficient.
This is not of course is not to argue that speculation and debt dont matter – to the contrary – simply to clear the way by removing fallacious Austrian diversions.
Can though am argument survive that the harmful effects of the Credit Cycle be diminished through 100% reserve money, as Fisher advocated? Ill tackle that in a future article.
 Lords Committee, Report, 1819, pp. 192-93.
 An illustration of Mr. Joplin’s views on currency, 1825, p.28
 Views on the subject of corn and currency, 1826, p.35.
 Studies in the Theory of International Trade New York: Harper and Brothers Publishers Viner IV.32
“r. Hayek on Money and Capital,” in The Economic Journal, Vol. 42, No. 1, March, 1932, pp. 42-53