Is State (outside) money a liability, and if so to whom?

The title of this post sums up a key problem and set of arguments in economics.  If all money is credit then it must be a liability.  Similarly if the price level is set by how much money is ‘backed’ rather than by its quantity then that backing must be in a liability form.  But it is less than obvious to whom that liability is owed in terms of central bank balance sheets especially when a state permits  a central bank to either increase them through an asset swap (QE), debt monetarisation or helicopter money.  The ability to alter the balance sheet in this manner ‘at the flick of a switch’ leads some – such as Eric Longeron, Wilhelm Buiter, Karl Whelan and  Paul De Grauwe, in various forms, to deny that state money is a liability.  There are also those like Anthony Werner who argue that money is endogenous (creditary) but that Central Bank (indeed for him all bank) balance sheets are not liability constrained- which again seems to redefine what is meant by ‘creditary’.  The implication that treating it as a liability of the Central Bank is just an ‘accounting convention’, that just can be ignored, as Ricardo Reis and others argue.

This is troubling for anyone concerned with the foundations of economic theory. It is like in physics discovering a solar system where gravity is no longer a constraint.  Often economists hold views which are not entirely consistent on this point.  For example many hold that all money is credit but that state money is not a liability in a consolidated central bank/government balance sheet.  This begs the question debt-liability to whom? Similarly there are those who deny that state money is a liability but deny monetarism, without noticing the inconsistency in terms of how the price of money is set.

Monetary theory has been somewhat in confusion since the mid 50s-60s, when the Samulelson treatment of the money multiplier (in the first edition of his textbook), Freidman and Swatz’s Monetary History of the United States (which set out the concept of an exogenous monetary base) and Gurley and Edward S. Shaw published Money in a Theory of Finance in 1960 (which first and confusing defined inside’ and ‘outside’ money).

The Gurley and Shaw argument

The economy has fiat money—an intrinsically useless asset with no backing whatsoever—that is generally accepted as a means of payment. A monetary authority or “government” has the monopoly to issue this asset. [Outside money]

However they simply asserted that agents would wish to hold this, Patinkin said this begged the question as to why?

These three texts together set the demonstrably false exogenous theory of money and loanable funds theory on a pedestal.  Not having a consistent global theory on these issues is a block to economics.  Without it we cannot fully understand the issues around Central Bank balance sheets, an issue central to most current public policy debates.

Readers of this blog will now that I have strongly argued that economics must be rebuilt around the accounting constraints of capitalism– balance sheet economics -, that these constraints are underlying laws of the economics system not legal conventions.  Whether or not conventions align with them is just contingent however if account don’t match them that are likely to lead to false economic decisions with real consequences.  From this perspective the resolution to such theoretical debates is straightforward in method but not always easy in practice, it is to determine what the accounting mistake is.

From the balance sheet perspective a liability will be held by an economic agent if it renders an economic service.  So what is the nature of that service?

A central bank expands its balance sheet in several ways but if state created money (specifically – in addition to bank created money) is to circulate then private banks must in turn expand their balance sheets.  The causation mechanism here is often confused.  And the text books (since Samuelson) show it reversed.  It is not that banks are granted excess reserves which are then ‘leant’, banks are profit constrained (and because of this capital constrained) not reserve constrained. State money is created through additional government spending from enlarged balance created at the central bank.  This leads to enlarged private bank balances and a regulatory constraint for additional reserves (held at the central bank) and for additional capital. Also it is not the creation of private bank central bank reserves (central bank liabilities), which help this circulation but excess reserves (1-private bank central bank reserves) the additional delta of state money (units of account) which adds to the unit of account and is held in privately held bank reserves.  More controversially I would argue that where there is a slow rate of turnover of any account the ‘goldsmith trick’ of fractional reserves can be applied whereby a firm temporarily use units of account of its customers to invest and then ensure the rate of payback to those accounts in totality does not exceed the rate of withdrawal (a bank run). These additional reserve ripple through the economy is a process of geometrical contraction.  Indeed this is the multiplier effect.  This is controversial but it is additional not central to this narrative.  There are those that hold this only holds in an age of commodity money not credit money, however providing anything can be counted and there is differential liquidity (liquidity transformation) between the reserve and the investment account then mathematically it must hold.  It can also be empirically observed in many financial sectors including paper gold firms, insurance (Warren Buffet is an expert in its use) and where retail and investment banks are held together. For a description of this theory see Werner here and here (though I disagree with his conclusion) or Humpreys.

Whilst bank created money not needed is used to pay loans and destroyed (reflux) this operates in a different way for state money.  Either it returns to bank reserves and acts as a permanent expansion of the unit of account it refluxes to other currencies and stays there.

Money therefore has a hierarchy in terms of its ability to make payments and circulates through the clearance of balances. The existence of taxes is one of the factors which places state money at the top of this hierarchy of money. Indeed ‘moneyness’ can be thought of as one a two dimensional access in terms of its ability to settle claims with a bank, with a clearing house, with a central bank and with the state, and the other of its liquidity.

The fact that money must be used to pay taxes to the sovereign forces acceptance of the unit of account -as chartalism argues.  But this is a necessary but not sufficient argument to determine the economic service that outside (state) money performs and hence whether it is a liability and to whom.  Certainly the level of taxes demanded is one often reasons holding money in the unit of account (as opposed to an arbitrary (such as a crypto currency) or non sovereign bank account) but it is not the only reason to hold money in the unit of account.

A simple model.  The exchange rate between a non sovereign bank account (which can be spent anywhere) and  a sovereign currency bank account is fixed.  The difference then between the value of one currency and the other being the liquidity services of holding money in that sovereign currency minus the cost of carry of that currency.  Strictly speaking it is the NPV of the ratio of these two factors that matters  but assuming that the non sovereign currency (which for sake of argument we will call the Bancor) has zero liquidity services and the cost of carry of both is zero this cancels and simplifies to the liquidity services of the sovereign currency.

Lets be strict about definitions.  By demand for liquidity services I don’t mean demand for money (which is infinite) or the demand for credit (which is rational leverage) but the opportunity cost of holding money in a sovereign unit of account. It is related to but not quite the same as liquidity preference.   Liquidity preference is a derived variable or both stock (wealth) and flow(income) components modulated by liquidity services.  Liquidity services are the inverse of the ‘hot potato’ ness of money, indeed if the unit of account were 100% hot it would never be held.

If liquidity services are positive then economic agents will have a demand to hold a an asset nominated in that unit of account, a liability to an issuer.

Historically the liability nature of central bank monies was clear.  Central Banks were privately held companies (some still are in whole or part) with liabilities to other including its own equity holders.  The liability nature is less clear in modern money, even though notes in England still retain the historical wording ‘I promise to pay the bearer’, especially with consolidated accounts, but is still there. If there is an economic service offered then a commodity (including money) has a positive economic value and agents will hold it as a liability.  Money is a commodity as well as being creditary (both controversial points) but its production is not costless.  A state has to persuade agents to hold that liability and the actions necessary to do so always has an opportunity cost in terms of real resources.

Now let us expand this model a little by assuming a small economic growth between two points in time generated by some technological innovation improving productivity.  This will create a demand for credit.  It will also produce a demand for the unit of account in which the good is produced. If wealth is held in other currencies then holders of the sovereign currency is which the good is produced will demand a profit.

As a simplifying assumption, which we relax later, let us assume an even distribution of creditors and debtors so that economic agents are net neutral about the effects of inflation/deflation on their wealth holdings.  In this case the sole cause in the price shift of the unit of account will come from changes in the relative price of the sovereign currency to the Bancor.  Let us say the innovation increases economic growth by 5% but the new good is only sold and traded in the country of the sovereign unit of account. In other words that country has a comparative advantage to the rest of the world.   This results in the innovation being deflationary (good deflation if you will).  A currency deflating relative to another will grow stronger. There will be arbitrage.  Holders of money will want it held in that unit of account and pay for liquidity services in that unit of account.  The producer of that unit of account is therefore able to extract seignorage from producing that currency.  That seignorage can take many forms but in the most demanded currency can even take the form of negative interest rates.

Now lets add a second sovereign currency, this time one without the comparative advantage.  Given its competitor currencies are deflating it must charge higher interest rates to compensate.

Holders of money in the innovating country will then not wish to move money out of the country.

But we have an innovation trap the national growth is completely vitiated by currency deflation.  In a global economy the holders of the capital of the innovating companies are likely to reside abroad whilst foreign goods and services will inflate. Hence the purchasing power of labour decreases and indeed labor may emigrate where it can.  With deflation banks are less willing to expand credit money as according to the fisher equation returns must be much higher for them to make a profit. For the purchasing power of labour to remain constant, for money not to be hoarded in the innovating country in the expectation of deflation and  for the lending power of banks to remain constant the unit of account must be expanded by the rate of growth of the economy. In other words the net state money destruction (taxation) and creation (public spending) must be positive, in surplus, equal to the rate of growth to maintain the price level.

In this model currency holders will maintain their wealth in their sovereign currency if the sovereign ‘backs’ the currency to the extent they maintain their wealth levels comparative to other currencies in which their wealth is held.  This it does through issuing gilts/treasuries.  The sale and purchase of these is secondary however.  The key is the net government deficit – supply creating state money – must equal the net additional demand  for liquidity services in that currency – demand.  Otherwise that currency will have a net appreciation or drain.

This concept is similar in many ways to the classical price-specie-flow mechanism, although here we are talking about a neutral ‘standard currency’ – neutral between all purchasing power rates internationally – a bancor.  In value theory terms this ‘standard currency’ fulfills the same role as Sraffas ‘standard commodity’ in his autarkical system.  It is ironical that this idea completely undermines the monetarist quantity theory in favour of the backing theory.

As a result of this exploration we can come to a number of conclusions:

• Money is anywhere and everywhere creditary and a commodity and backed, but a commodity of a very different kind, it is an inherently valueness token which has value not for what it is but for what it does, which is to provide liquidity services.
• Money is never neutral let alone superneutral, the distribution of wealth and of debtors/creditors affects the demand for liquidity services in any one unit of account over another.
• There are no black holes in economics where the fundamental equation of accounting does  not apply – therefore a general economic theory is possible.  Anyone claiming otherwise is a charlatan who requires some kind of mysterious bootstrap economic force.
• Monetary and fiscal policy are inherently bound in floating currencies.  QE for example helps relax the constraints on fiscal policy.
• Without fiscal deficits you enter into a global deflationary spiral and currency war.
• No every country can deflate together -its a zero sum game, indeed negative sum because of the impacts on nominal values of debts.
• Assets and liabilities have real meaning on Central Bank balance sheets, and Central Banks have an ability to create price neutral money (charter value) just like private banks (I will pursue this is a more academic coming article with T accounts) -see for example Keen, Bezemer, Hudson and Graselli.
• The fiscal theory of the price level is flawed, the real constraint is a solvency constraint on Central Banks whilst maintaining a neutral price level (what Scott Fullwiler calls an inflation constraint) – sovereign currency issuers need never go insolvent and can always create liabilities s providing these have positive liquidity services  – but the costs of doing so to the currency may be unacceptable.  Emerging economies may prefer instead to default and often have.  States can add capital to CBs to enable them to issue more liabilities, but doing so always has economic consequences. All Central Banks cannot expand their balance sheets infinitely and together at the same time, this is a fallacy of aggregation, the largest most attractive currencies will always prevail.
• Excess reserves matter, but for precisely the opposite reason found in graduate textbooks.
• The above treatment of seignorage is simplified due to the sheer variety of central bank operations, but the underlying issues and drivers remain the same.
• Value theory and monetary theory are connected in fundamental and deep ways – it was no coincidence that the Theory of Comparative Advantage was set out in the LTV.
• The value of liquidity services helps explain the proportion of inside to outside money in existence at any one time.

In the more academic piece I will do a full list of references.