Tag Archives: endogenous money

We Don’t Need No Helicopters … Hey! Economists! Leave Fiscal Policy Alone

A lot has been written on helicopter money recently. Most of them bad with a few exceptions such as one by JKH.

In my opinion, the main reason economists come up with stories such as “helicopter money” etc. is that it is difficult in standard economic theory to introduce money.

Few quotes from Mervyn King’s book The End of Alchemy: Money, Banking, and the Future of the Global Economy:

But my experience at the Bank also revealed the inadequacies of the ‘models’ – whether verbal descriptions or mathematical equations  – used by economists to explain swings in total spending and production. In particular such models say nothing about the importance of money and banks and the panoply of financial markets that feature prominently in newspapers and on our television screens. Is there a fundamental weakness in the intellectual economic framework underpinning contemporary thinking? [p 7]

For over two centuries, economists have struggled to provide a rigorous theoretical basis for the role of money, and have largely failed. It is a striking fact that as as economics has become more and more sophisticated, it has had less and less to say about money… As the emininent Cambridge economist, and late Professor Frank Hahn, wrote: ‘the most serious challenge that the existence of money poses to the theorist is this: the best developed model of the economy cannot find room for it’.

Why is modern economics unable to explain why money exists? It is the result of a particular view of competitive markets. Adam Smith’s ‘invisible hand’ …

… Money has no place in an economy with the grand auction. [pp 78-80]

But the ex-Bank of England governor perhaps never worked with stock flow consistent models. The advantage of these models is that what money is and how it is created is central to the question of how economies work. The framework used in stock flow consistent models is not new exactly. What’s new in stock-flow consistent models is the behavioural analysis on top of the existing framework the system of national accounts and flow of funds. As Morris Copeland, who formulated the flow of funds accounts of the U.S. economy said:

The subject of money, credit and moneyflows is a highly technical one, but it is also one that has a wide popular appeal. For centuries it has attracted quacks as well as serious students, and there has too often been difficulty in distinguishing a widely held popular belief from a completely formulated and tested scientific hypothesis.

I have said that the subject of money and moneyflows lends itself to a social accounting approach. Let me go one step farther. I am convinced that only with such an approach will economists be able to rid this subject of the quackery and misconceptions that have hitherto been prevalent in it.

– Morris Copeland, Social Accounting For Moneyflows in Flow-of-Funds Analysis: A Handbook for Practitioners (1996) [article originally published in 1949]

So what do we mean by helicopter money and it is really needed or useful? For that we need to go into a bit into some behavioural equations in stock-flow consistent models. One way is to use a somewhat simplified notation from Tobin’s nobel prize lecture Money and Finance in the Macroeconomic Process. In Tobin’s analysis, the government’s fiscal deficit is financed by high-powered money and government bonds:

GT = ΔH + ΔB

ΔH = γH·(G – T)

ΔB = γB·(G – T)

 γ+ γ= 1

0 ≤  γH, γB  ≤ 1

So the deficit is financed by “high-powered money” (H) and government bonds (B) in proportion γand γB

Now it is important to go into a bit of technicalities. Prior to 2008, central banks implemented monetary policy by a corridor system. After 2008, when the financial system needed to be rescued and later when central banks started the large scale asset purchase program (“QE”), central banks shifted to a floor system.

Although economics textbooks keep claiming that the central bank “controls the money supply”, in reality they are just setting interest rates.

In the corridor system, there are three important rates:

  1. The deposit rate: The rate at which central banks pay interest on banks’ deposits (reserves) with them,
  2. The target rate: The rate which the central bank is targeting, and is typically the rate at which banks borrow from each other, overnight, at the end of the day.
  3. The lending rate: The rate at which the central bank will lend to banks overnight.

There are many complications but the above is for simplicity. Typically the target rate is mid-way between the lower (deposit rate) and the higher (lending rate).

In the floor system, the government and the central bank cannot set the overnight at the target rate if the central bank doesn’t supply as much reserves as demanded by banks. Else the interest rate will fall to the deposit rate or rise to the lending rate. In a system with a “reserve-requirement”, banks will need an amount of reserves deposited at the central bank equal to a fraction of deposits of non-banks at banks.


H = ρ·M

where M is deposits of non-banks at banks and ρ is the reserve requirement. In stock-flow consistent models, is endogenous and cannot be set by the central bank. Hence is also endogenous.

In the floor system, the target rate is the rate at which the central bank pays interest on deposits. Hence the name “floor”. There are some additional complications for the Eurosystem, but let’s not go into that and work in this simplification.

In the floor system, the central bank and the government can decide the proportions in which deficit is financed between high powered money  and government bonds. However since deposits are endogenous the relation between high powered money and deposits no longer holds.

In short,

In a corridor system, γand γB are endogenous, M is endogenous and H = ρ·M. In a floor system, γand γB can be made exogenous, M is endogenous and H ≠ ρ·M. is not controlled by the central bank or the government in either cases and is determined by asset allocation decisions of the non-bank sector.

Of course, the government deficit Gitself is endogenous and we should treat the government expenditure G and the tax-rates θ as exogenous not the deficit itself.

So we can give some meaning to “helicopter money”. It’s when the central bank is implementing monetary policy by a floor system and γand γB are exogenous.

But this doesn’t end there. there are people such as Ben Bernanke who have even proposed that the central bank credit government’s account with some amount and let it spend. So this introduces a new variable and let’s call it Gcb.

So we have a corridor system with variables G and θ versus a floor system with variables G’G’cbθ,  γ’and γ’B

The question then is how is the latter more superior. Surely the output or GDP of an economy is different in the two cases. However people constantly arguing the case for “helicopter money” are in the illusion that the latter case is somewhat superior. Why for example isn’t the vanilla case of a corridor system with higher government expenditure worse than “helicopter money”.

Also it effectively reduces to a fiscal expansion combined with a large scale asset purchase program of the central bank (“QE”). I described QE’s effect here. Roughly it works by a wealth effect on output with some effect on investment via asset allocation.

To summarize, the effect on output by these crazy ways can be achieved by a higher fiscal expansion. There’s hardly a need to bring in helicopters. Some defenders say that it is faster but that just sounds like an excuse to not educate policymakers.

Money Stock Determination

The recent Bank of England quarterly bulletin has interested blogosphere into what goes on to determine the stock of money.

Money can mean various things and here I restrict to the the monetary aggregates as defined by central banks – as in the referred publication. But whoever is interested in “money creation” also becomes interested in the creation of assets and liabilities, so the right question is more general.

As I had pointed out in my previous post, the Bank of England articles seriously ignore the role of fiscal policy. Winterspeak also mentions this.

So what is the answer? In my view the most systematic way of saying this via Tobin’s theory asset allocation, improved drastically by stock-flow consistent models of Godley and Lavoie.

Also there are two things – influence and determination. For example, something can have an influence on the stock of money but may not determine it.

Since economies are highly dynamic it is not easy to answer this in a single sentence but it can perhaps be said that fiscal policy, private expenditure and QE influence the stock of money but it is ultimately determined by the holders of wealth.

Of course since people generally have a Monetarist intuition, the right notion that fiscal policy, private expenditure and QE influencing the determination of the stock of money is incorrectly taken by people to mean that QE has the same effect as a fiscal expansion. Which of course QE does not.

First take private expenditure. Since we know that “loans create deposits” it can be suspected that bank credit has an influence on money. Of course this process is more dynamic as the expenditure has its own multiplier effect (not to be confused with the money mulplier!) on output and income. But bank credit determining the stock of money is stretching too much. For example, while a bank makes a house loan and creates deposits in the process, the process of securitization reduces the stock of money as ultimate buyers of the securized products exchange money with the mortgage-backed securities (MBS). And of course there’s the reflux mechanism via which economic units may reduce their debts toward the banking system.

Now take fiscal policy.  Like private expenditure, government expenditure and taxes also influence the level of aggregate demand. This has an influential effect on credit creation via effect of increased output and income on private expenditure and via the process highlighted in the previous paragraph this has an influence on the determination of the stock of money.

Also, while economic units are earning and making decisions on spending, they are also accumulating financial and non-financial assets. So they have a preference on how much of their wealth they allocate into each asset. A Monetarist would talk of an excess supply of money and this raising prices of goods and services and bringing the demand and supply of money into equivalence. But there is no need for this from an endogenous money perspective. One can have the equivalence brought about by adjustments of prices of financial assets and also adjustment of quantities of assets and liabilities held by various economic units such as banks. This is where the importance of the work of Tobin’s theory of asset allocation comes in.

Now let’s discuss QE. Large scale purchases of financial assets by the central bank – although influences the stock of money, doesn’t determine it. Also QE doesn’t have a direct influence on aggregate demand like private or public expenditure. It has indirect effects via raising prices of financial assets (which can be described by Tobin’s theory of asset allocation) and inducing capital gains and a wealth effect on consumption. The Monetarist intuition highly exaggerates the effect.

The point of my writing the post was to show that fiscal policy has a strong effect on influencing the stock of money. This happens via the strong effect of fiscal policy on output and income inducing private expenditure. Of course private expenditure needn’t be only induced and has an autonomous nature as well, so both fiscal policy and private expenditure have an effect. The effect is via a rise in output and income and this leading to a rise in wealth and economic units allocating a fraction of their increased wealth into ‘money’ (as in currency notes and deposits).

So Winterspeak is right in pointing out the incorrect statement of the Bank of England paper:

The amount of money created in the economy ultimately depends on the monetary policy of the central bank. In normal times, this is carried out by setting interest rates.

The above quote is suggestive of a very strong influence of interest rates on private expenditure, ignores the autonomous nature of private expenditure and the role of fiscal policy.

Nick Rowe defends textbook economics over his blog and suggests some influence of QE on prices via the Monetarist hot potato process where there is an excess supply of money and via a non-equilibrium process leads to a rise in prices of good and services! But in this he mixes asset allocation decisions with expenditure decisions, as if the two can be treated as the same. In the market for goods and services, producers set the price based on costs and their markups. So it is hard to see the influence. The supplies and demands of assets are actually brought into equivalence in the financial markets rather than the market for goods and services. He may have a point but the degree to which this has an effect is low. So holders of wealth may allocate some of their portfolio into commodity funds (after having sold their bonds to the central bank) which may buy commodities in exchanges and expectations due to a price rise and speculation and myths may cause a price rise but this is quite different from his suggested dynamics. Or it may have an effect via a depreciation of the currency and change in the consumer price index due to a change in prices of foreign goods.  The question then is to what extent do what economists stress are important are actually important. 

Needless to say, the usual story from money to other things is misleading. The point however is that “how money is created” is a good starting point to understand macroeconomics.

Bank Of England On Money Creation

In the natural sciences, controversies are settled in a few months, or at a time of crisis, in a year or two, but in the social so-called sciences, absurd misunderstandings can continue for sixty or a hundred years without being cleared up.

– Joan Robinson, 1981 (1979), What Are The Questions And Other Essays – Further Contributions To Modern Economics, M.E. Sharpe

The latest Bank of England Quarterly Bulletin (2014 Q1) will be released on the 14th. It has pre-released two articles which go into money creation and the myths associated with it. 

The page is here. The second article Money creation in the modern economy may interest you more but the first is also readable.

Interestingly, the second pape refers to Post-Keynesians : Tom Palley’s 1996 book , Basil Moore’s 1988 book, a JPKE paper by Peter Howells and a 1981 paper by Nicholas Kaldor and J. Trevithick which discusses the reflux mechanism (reprinted in Kaldor’s Collected Economic Essays, Vol. 9). It also refers to James Tobin’s 1963 paper Commercial Banks As Creators Of “Money”. 

One negative is the omission of fiscal policy from the discussion altogether and emphasising monetary policy. This underplay of fiscal policy and overemphasis of monetary policy is one deep bias of the profession. The paper also has a slightly different emphasis on what determines the quantity of lending than emphasized by Post-Keynesians but I won’t go into it now. Still the page is worth a look. 

The full bulletin will be available on this page in a couple of days.

Update: The webpage for the full quarterly report is now available and it is here: Quarterly Bulletin 2014 Q1

James Tobin Turns In His Grave

When I was a boy of 14, my father was so ignorant I could hardly stand to have the old man around. But when I got to be 21, I was astonished at how much the old man had learned in seven years.

– attributed to Mark Twain, Reader’s Digest, September 1939.

Steve Keen has a blog post The Getting Of Wisdom in which he compares Tobin’s view in his 1963 paper Commerical Banks As Creators Of “Money” to his 1982 paper The Commercial Banking Firm: A Simple Model and finds Tobin has so many things (i.e., “loans create deposits” in 1963 to “loans create deposits” in 1982) in those years!

According to Keen:

The difference between the Old and New Tobin is as stark as that between the Old and New Testament. Not only is there an emphasis on the uniqueness of banks in that 1982 paper, Tobin also makes copious use of T-accounts and double-entry bookkeeping to explain why banks do matter. So just as the Testament message moved from “An eye for an eye” to “Turn the other cheek”, Tobin moved from “banks don’t matter and the belief that banks create money is a shibboleth”, to “banks are crucial to macroeconomics and they can and do create money”.

And whereas Tobin the Younger imagined that newly created bank money could be taken out of the system in a form other than bank deposits or cash, Tobin the Elder realizes that those are the only two options at the systemic level. Individuals might get out of bank deposits into (say) gold, but to do so they transfer money from their deposits in one bank into the deposits of the gold dealer in another bank. The only way for money not to be held in a bank is for it to be converted into some other kind of asset that is not a bank liability first. The only candidate here is cash—notes and coins—which you can insist on when you make a withdrawal (you might insist on gold instead, but a bank is under no obligation to deliver it in response to your withdrawal).

(italics in original, boldening mine)

Keen obviously is wrong when he says “the only way for money to be not held … ” because he first he misses the reflux mechanism where money can be extinguished by reducing indebtedness to the banking system. He obviously knows what “reflux” is but nonetheless misses it. Among other things he misses is that the bank can induce the non-banking public to shift in and out of money (or in the opposite case accommodate the non-banking public’s desire to change its portfolio) by changing bid/asks of markets they are dealers in such as government bond prices. So there is a price adjustment in the financial markets as well (leading to changes in deposits)

[Of course Keen attributes it to Tobin but looking at his agreeing tone, looks like it his own claim].

Others examples include but not limited to – a shift of deposits abroad with accommodative transactions which do not bring back the deposits to the original level, a “Treasury Supplementary Financing Operation” type of operation by the government Treasury which reduces the deposits in existence etc.

But more importantly he misses the mechanism which Tobin highlighted in his 1963 paper in which non-bank financials can compete with banks in the loan markets by taking away some borrowers and this leading to a fall in deposits. And this way, it is important not only for a single bank but for the banking system to induce depositors to bank with them.

Another thing which Post-Keynesians (some, not all of course) sometimes do is to think erroneously that bank borrowing (for purchases of goods and services at any rate) adds to demand and that non-bank borrowing/lending is demand-neutral and reading Keen’s post it seems he is close to saying something of the sort. And in his models, borrowing for purchases of financial assets also adds to “aggregate demand”.

I had written the following and am repeating it again about how a non-bank induces a borrower to borrow from it (who then extinguishes his loan from banks to become a borrower from the non-bank financial) and how this reduces the banking system’s balance sheet and hence deposits in the process – something Keen can’t see.

The following example is given in Tobin’s 1963 paper with my own numbers.

Start with a bank with initial balance sheet of 100 units. I will neglect capital and other liabilities to keep things clean so if you are not comfortable you can always change the liabilities side by reducing deposits by say 10 and replacing it with other liabilities. Also I call NBFIs’ liabilities “shares” and this is more like money-market mutual fund shares and shouldn’t be confused with stock-market shares.

t = 0


Assets: Loans = 100
Liabilities: Deposits = 100

Non-Financial Private Sector

Assets: Deposits = 100
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Deposits = 0
Liabilities: Shares = 0

t = 1

At t = 1, let us say NBFIs attract 10 units of deposits from bank depositors. So the balance sheets will look like:


Assets: Loans = 100
Liabilities: Deposits = 100

Non-Financial Private Sector

Assets: Deposits = 90, Shares = 10
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Deposits = 10
Liabilities: Shares = 10

t = 2

At t = 2, someone extinguishes his/her/its loan to the banking system by 10 unit. So,


Assets: Loans = 90
Liabilities: Deposits = 90

Non-Financial Private Sector

Assets: Deposits = 80, Shares = 10
Liabilities: Loans = 90

Non-bank Financial Institutions

Assets: Deposits = 10
Liabilities: Shares = 10

t = 3

At t = 3, someone borrows 10 units from NBFIs. So,


Assets: Loans = 90
Liabilities: Deposits = 90

Non-Financial Private Sector

Assets: Deposits = 90, Shares = 10
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Loans = 10
Liabilities: Shares = 10

NBFIs who had 10 units of deposits no longer have it because they have lent 10 units which involves transfer of deposits. The net result at the end is that banks have lost a share lending of 10 units out of the initial 100 to non-banks and also deposits worth 10 units.

This of course can go on and it is in the interest of banks to prevent this from happening and induce the public to bank with them. In Tobin’s asset allocation theory, asset demands are dependent on the portfolio preference parameter and also the interest rate paid on the asset (or expected returns in general). So putting up interest rates on deposits would partly prevent this shift to non-bank financial intermediaries.

So in trying to show something, Keen’s effort turns counter-productive because his claim:

The only way for money not to be held in a bank is for it to be converted into some other kind of asset that is not a bank liability first.

turns out to be wrong and misleading.

He misses out Tobin’s insight that banks individually and collectively have to induce the non-banking public to hold deposits with them in his 1963 paper which the simple “loans create deposits” phrase does not highlight and how the banking system’s deposits can reduce due to competition from non-bank financials.

Plus there are more Tobinesque mechanisms (via his asset allocation theory) as I highlight in my previous posts James Tobin, Banking And The Widow’s Cruse and Holier Than Tobin? in which price changes of financial assets leads to a change in the stock of money which Keen is not aware of.

Instead his post has basic errors in monetary economics.

James Tobin, Banking And The Widow’s Cruse

There is good discussion in the blogosphere on James Tobin’s 1963 paper Commerical Banks As Creators Of “Money” – also mentioned in my previous post Holier Than Tobin?

This blog post is an attempt to present Tobin’s ideas from the paper in a more simplistic way.

One of  Tobin’s points is a critique of the notion that since loans create deposits, it makes banks special as compared to non-bank financial institutions and the over-emphasis on this point by many.

Tobin goes on to show how this is misleading. The fact that a non-banking financial institutions don’t simply credit shares like banks is not too important.

From the viewpoint of a single bank, while loans make deposits, the deposits can “fly out” to another bank and hence the bank is limited by its deposit raising ability. In general, a bank can fund itself by using other things – not just  deposits – so a bank will need to fund itself. It is sometimes said that “banks lend first and look for deposits later” but this is a bit misleading because while it is true in general, it is in the confident knowledge that the funding will be available at a not so costly rate. If the bank fears or the whole banking system fears a funding crisis, then lending will be curtailed.

It is true that the bank can fund itself from the central bank but even this is not available for unlimited amount. It has to provide collateral to the central bank which is limited. Usually these are marketable securities and not loans provided to the private sector and the amount of marketable securities is a small fraction of banks’ balance sheet.

At a macro level however, deposits leaving a bank may move to another bank so one may conclude that the banking sector as a whole collectively possesses a Widow’s Cruse.

Tobin however goes on to show how the presence of non-banking financial institutions (NBFIs) presents problems for such a view – by hook or crook, the banking system has to induce the non-banking private sector to hold deposits with them than depositing it with NBFIs. In general flight of deposits abroad is also important. This comes at a cost – the easiest to think of is the interest rate paid on deposits but one can also think of other things such as advertising costs etc.

In the following I show how this happens and how the banking system’s balance sheet can shrink because of flight of deposits to NBFIs who can take away banks’ market share. The fact that loans create deposits is not so important as is emphasized many times. Even though non-banks cannot simply credit the “share” account  doesn’t mean much. They can keep attracting deposits from banks and lend.

So let us take a simple example: start with a bank with initial balance sheet of 100 units. I will neglect capital and other liabilities to keep things clean so if you are not comfortable you can always change the liabilities side by reducing deposits by say 10 and replacing it with other liabilities. Also I call NBFIs’ liabilities “shares” and this is more like money-market mutual fund shares and shouldn’t be confused with stock-market shares.

t = 0


Assets: Loans = 100
Liabilities: Deposits = 100

Non-Financial Private Sector

Assets: Deposits = 100
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Deposits = 0
Liabilities: Shares = 0

t = 1

At t = 1, let us say NBFIs attract 10 units of deposits from bank depositors. So the balance sheets will look like:


Assets: Loans = 100
Liabilities: Deposits = 100

Non-Financial Private Sector

Assets: Deposits = 90, Shares = 10
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Deposits = 10
Liabilities: Shares = 10

t = 2

At t = 2, someone extinguishes his/her/its loan to the banking system by 10 unit. So,


Assets: Loans = 90
Liabilities: Deposits = 90

Non-Financial Private Sector

Assets: Deposits = 80, Shares = 10
Liabilities: Loans = 90

Non-bank Financial Institutions

Assets: Deposits = 10
Liabilities: Shares = 10

t = 3

At t = 3, someone borrows 10 units from NBFIs. So,


Assets: Loans = 90
Liabilities: Deposits = 90

Non-Financial Private Sector

Assets: Deposits = 90, Shares = 10
Liabilities: Loans = 100

Non-bank Financial Institutions

Assets: Loans = 10
Liabilities: Shares = 10

NBFIs who had 10 units of deposits no longer have it because they have lent 10 units which involves transfer of deposits. The net result at the end is that banks have lost a share of 10 units out of the initial 100 to non-banks and also deposits worth 10 units.

This of course can go on and it is in the interest of banks to prevent this from happening and induce the public to bank with them. In Tobin’s asset allocation theory, asset demands are dependent on the portfolio preference parameter and also the interest rate paid on the asset (or expected returns in general). So putting up interest rates on deposits would prevent this shift to non-bank financial intermediaries.

Tobin would say that “at this point the widow’s cruse has run dry”. Perhaps there is an over-emphasis on this but I leave it to the reader to decide.

One thing Tobin didn’t emphasise is the role of effective demand. I would imagine he would explain why lending doesn’t explode by using some neoclassical marginal curves instead of the Post-Keynesian answer.

Holier Than Tobin?

It sometimes happens that important insights of great contributors to an academic field are missed. One of the most important things in Monetary Economics is Tobin’s asset allocation theory which although is well known is sometimes poorly understood.

James TobinJames Tobin (Source: Econometric Theory)

But sometimes a holier-than-thou attitude can lead one to miss an important and insightful aspect of a work.

The blogger Winterspeak – well aware of some of Tobin’s work such as his paper Commercial Banks As Creators Of “Money” from  1963 has written as post A Bank is not a Financial Intermediary and concludes that

… This then is the conceptual fallacy at the heart of academic macro and what it thinks about banks, and it goes at least all the way back to 1963.

Winterspeak is stuck on a quote from Tobin-Brainard paper (1963) which says:

…the essential function of financial intermediaries, including commercial banks, is to satisfy simultaneously the portfolio preferences of two types of individuals or firms. On one side are borrowers, who wish to expand their holdings of real assets… On the other side are lenders who wish to hold part or all of their net worth in assets of stable money value with negligible risk of default.

This is also repeated in Tobin’s Commercial Banks As Creators Of “Money” which obviously states explicitly that loans create deposits and that the money mutliplier view is highly inaccurate:

According to the ‘new view’, the essential function of financial intermediaries, including commercial banks, is to satisfy simultaneously the portfolio preferences of two types of individuals or firms. On one side are borrowers, who wish to expand their holdings of real assets – inventories, residential real estate, productive plant and equipment, etc. – beyond the limits of their own net worth. On the other side are lenders, who wish to hold part or all of their net worth in assets of stable money value with negligible risk of default. The assets of financial intermediaries are obligations of the borrowers – promissory notes, bonds, mortgages. The liability of financial intermediaries are the assets of the lenders – bank deposits, insurance policies, pension rights.

Winterspeak is adamant about the usage of the phrase “intermediary” and that since banks create deposits out of thin air, they shouldn’t really be called intermediary and that Tobin’s views are equivalent to the loanable funds view. For the first part – maybe but Winterspeak seems to crucially miss out the mediating role played by banks in the portfolio allocation decisions of wealth owners such as households. See my comments in that blog.

First it is crystal clear that Tobin knows loans create deposits. Second, he presents a “new view” in which the distinction between “money” and “bonds” is blurred and this led him subsequently to his asset allocation theory. It is true that Tobin’s model was far from complete and this was improved substantially by Wynne Godley, but nevertheless Tobin’s insights were wonderful and the work of a genius.

Perhaps I would have worded what Tobin wrote differently if I were to teach this but this is just a matter of emphasis.

Perhaps “the essential function” is better worded with “an essential function” so that the reader doesn’t take it to mean that the concept I will come to,  isn’t taken to mean “the only function” or “the most essential function”.

The mediation role played by banks is related to the super-version of “loans create deposits” – asset purchases by banks also create deposits.

So when a bank purchases say a government bond from a household (or a mutual fund selling in response to redemptions by a household), banks create more deposits in the process. In the opposite case, there is a destruction of deposits.

Now suppose for some reason – such as improved animal spirits of entrepreneurs, firms borrow more from banks and the expenditures transfers funds to households. Coincidentally – for different reasons – households wish to hold less of their wealth in deposits and more in bonds or other securities. There is now a discrepancy and it is reconciliated by banks standing ready to sell bonds to households. This decreases the stock of money (monetary aggregate) in existence so that there is no discrepancy at all. There is of course another way in which this may happen – i.e., by price changes (of financial secruities and not that of goods and services) bringing supplies equal to demand but this needs a full course on asset allocation theory discovered by nobody else than James Tobin himself!!

Of course there are other ways. There is the reflux mechanism and more complicated mechanisms involving asset allocation theory such as higher issuance of equities by production firms. In the reverse case when households desire to hold more of their wealth in deposits, firms may need to borrow more from banks so that the “supply” of money is equal to the “demand”.

In contrast there is the Monetarist hot potato process which mainly relies on prices changes of goods and services. In ideas such as the asset allocation theory including the mechanism of the mediating role of banks is a blow to the Monetarist hot potato.

Of course there is the notion of convenience lending – one favoured by Basil Moore – in which household volitionally hold bank deposits non-volitionally but it is too artificial.

This mediating role of banks (and not the most important if you like) is also endorsed by some Post-Keynesian authors such as Wynne Godley and Nicholas Kaldor.

In an article In his essay Keynes And The Management Of Real Income And Expenditure, (in Keynes And The Modern World, ed. George David Norman Worswick and James Anthony Trevithick, Cambridge University Press, 1983), Wynne Godley says (p 151):

Even though I am not going in detail into monetary mechanisms it is worth drawing attention to the fact that the commercial banks’ role, apart from creating credit to finance certain kinds of expenditure, is to mediate the non-bank private sector’s portfolio choice, given the income flows and the central authorities’ funding policy.

Nicky Kaldor’s The Scourge Of Monetarism (Oxford University Press, 1982) is more clearer than simply stating:

As it is, a highly developed banking system already provides such facilities on an ample scale, since it is prepared to accommodate the public’s changing demand between different types or financial assets by altering the composition of the banks’ assets or liabilities in a reverse direction. If the non-banking public wishes to switch its holding of gilts for interest-bearing bank deposits, the banks are ready to supply such deposits at the minimum of inconvenience, and at the same time to place their surplus funds into the gilts which were previously held by the public. Similarly the banks provide easy facilities to their customers for switching balances on current accounts into interest-bearing deposit accounts, or vice versa. Hence, while the annual increment in the total holding of financial assets of the private sector (considered as a whole) is nothing more than the mirror-image of the borrowing requirement of the public sector (in a closed economy at any rate), neither the Government nor the banks can determine how much of this increment will be held in the form of cash (meaning notes and current deposits) and how much in the near-equivalents to cash (such as interest-bearing demand deposits) or in various forms of public sector debt. Thus neither the Government nor the central bank can control how much or the total financial assets the public prefers to hold in the form of ‘money’ on one particular definition or another.

Again in 1997 in Money Finance And National Income Determination Wynne Godley repeats himself although criticising Tobin but nevertheless realising the importance of his work – this time writing an explicit model for the whole thing which incorporates Tobin’s ideas:

… I am saying that (within strict limits e.g. concerning credit-worthiness) banks respond passively to the needs of business for loans and to the asset allocation activities of households (as well as providing the means of payment).


It is true that PKE authors and bloggers do have a much better understanding of monetary matters than mainstream economists but in trying to emphasise this point, sometimes they miss out on important matters. There is no need to say (as Winterspeak says “Tobin … sees … [banks as] something which brings efficiency and eases friction between the actual lender and borrower.”) especially when Winterspeak doesn’t seem to understand the mediating role of banks in the portfolio allocation decisions of financial asset owners which really has less to do with any “friction”. Perhaps the word intermediary is not the best but it is a minor point. In fact the ideas of the 1960s and later are missed by younger ones.

Money And Shoes

… Now let me give you a ridiculous example to make the point. Don’t take it too seriously. Suppose that some statistician observes that over a long period of time there is a high association, a very good fit, between gross national product and the sales of, let us say, shoes. And then suppose someone comes along and says, “That’s a very good relationship. Therefore, if we want to control GNP, we ought to control production of shoes. So, henceforth, we’ll make shoes grow in production precisely at 4 percent per year, and that will make GNP do the same.” I don’t think you would have much confidence in drawing this second conclusion and policy recommendations from the observed empirical association.

Over the years, according to the monetarists, the Federal Reserve has been acting like the producers and sellers of shoes. That is, the Fed has been supplying money on demand from the economy instead of using the money supply to control the economy. The Fed has looked at the wrong targets and the wrong indicators. As a result, the Fed has allowed the supply of money to creep up when the demand for money rose as a result of expansion in business activity, and to fall when business activity has slacked off. This criticism implies that the supply of money has, in fact, not been an exogenously controlled variable over the period of observation. It has been an endogenous variable, responding to changes in economic conditions and credit market indicators via whatever response mechanism was built into the men in this room and their predecessors.

… Perhaps the monetarists will be sufficiently persuasive of the Federal Reserve and of Congressional committees to bring about, in the future, a controlled experiment in which the stock of money is actually an exogenous variable.

– James Tobin, 1969


  1. Tobin, James, “The Role of Money In National Economic Policy – A Panel Discussion,” in Controlling Monetary Aggregates. Boston: Federal Reserve Bank of Boston, 1969, pp. 21-24 (link)

William Dudley On Bank Lending

I came across this 2009 speech by William Dudley on banking where he rightly says:

… A related concern is the question of whether the Federal Reserve will be able to act quickly enough once it determines that it is time to raise rates. This concern reflects the view that the excess reserves sitting on banks’ balance sheets are essentially “dry tinder” that could quickly fuel excessive credit creation and put the Fed behind the curve in tightening monetary policy.

In terms of imagery, this concern seems compelling—the banks sitting on piles of money that could be used to extend credit on a moment’s notice. However, this reasoning ignores a very important point. Based on how monetary policy has been conducted for several decades, banks have always had the ability to expand credit whenever they like. They don’t need a pile of “dry tinder” in the form of excess reserves to do so. That is because the Federal Reserve has committed itself to supply sufficient reserves to keep the fed funds rate at its target. If banks want to expand credit and that drives up the demand for reserves, the Fed automatically meets that demand in its conduct of monetary policy. In terms of the ability to expand credit rapidly, it makes no difference whether the banks have lots of excess reserves or not. 

[emphasis: mine]


Lots of central bankers have nuanced views on this but mostly end up believing the money multiplier approach at some point and it is rare to see this.

Kaldor’s Reflux Mechanism

The recent debates of Post Keynesians with Neoclassical/New Keynesians has highlighted that the latter group continues to hold Monetarists’ intuitions. Somehow the exogeneity of money is difficult for them to get rid of, in spite of their statements and rhetoric that money is endogenous in their models.

So there is an excess of money in their models and this gets resolved by a series of buy and sell activities in the “market” (mixing up decisions of consumption and portfolio allocation) until a new “equilibrium” is reached where there is no excess money.

Two-Stage Decision

While the following may sound obvious, it is not to most economists. Keynes talked of a two-stage decision – how much households save out of their income and how they decide to allocate their wealth. In the incorrect “hot potato” model of Neoclassical economists and their New “Keynesian” cousins, these decisions get mixed up without any respect for the two-stage decision. It is as if there enters an excess money in the economy from somewhere and people may consume more (as if consumption is dependent on the amount of deposits and not on income) till prices rise to bring the demand for money equal to what has been “supplied” (presumably by the central bank).

In their book Monetary Economics, Wynne Godley and Marc Lavoie have this to say (p 103):

A key behavioural assumption made here, as well as in the chapters to follow, is that households make a twostage decision (Keynes 1936: 166). In the first step, households decide how much they will save out of their income. In the second step, households decide how they will allocate their wealth, including their newly acquired wealth. The two decisions are made within the same time frame in the model. However, the two decisions are distinct and of a hierarchical form. The consumption decision determines the size of the (expected) end-of-period stock of wealth; the portfolio decision determines the allocation of the (expected) stock of wealth. This behavioural hypothesis makes it easier to understand the sequential pattern of household decisions.

[Footnote]: In his simulation work, but not in his theoretical work, Tobin endorsed the sequential decision that has been proposed here: ‘In the current version of the model households have been depicted as first allocating income between consumption and savings and then making an independent allocation of the saving among the several assets’ (Backus et al. 1980: 273). Skott (1989: 57) is a concrete example where such a sequential process is not followed in a model that incorporates a Keynesian multiplier and portfolio choice.

Constant Money?

So even if one agrees that loans make deposits, there is still a question of deposits just being moved around and the (incorrect) intuition is that someone somewhere must be holding the deposit and hence similar to the hot potato effect. The error in this reasoning is the ignorance that repayment of loans extinguishes money (meant to be deposits in this context).

Nicholas Kaldor realized this Monetarist error early and had this to say

Given the fact that the demand for money represents a stable function of incomes (or expenditures), Friedman and his associates conclude that any increase in the supply of money, however brought about (for example, through open-market operations that lead to the substitution of cash for short-term government debt in the hands of discount houses or other financial institutions), will imply that the supply of money will exceed the demand at the prevailing level of incomes (people will “find themselves” with more money than they wish to hold). This defect, in their view, will be remedied, and can only be remedied, by an increase in expenditures that will raise incomes sufficiently to eliminate the excess of supply over the demand for money.

As a description of what happens in a modern economy, and as a piece of reasoning applied to situations where money consists of “credit money” brought about by the creation of public or private debt, this is a fallacious piece of reasoning. It is an illegitimate application of the original propositions of the quantity theory of money, which (by the theory’s originators at any rate) were applied to situations in which money consisted of commodities, such as gold or silver, where the total quantity in existence could be regarded as exogenously given at any one time as a heritage of the past; and where sudden and unexpected increases in supply could occur (such as those following the Spanish conquest of Mexico), the absorption of which necessitated a fall in the value of the money commodity relative to other commodities. Until that happened, someone was always holding more gold (or silver) than he desired, and since all the gold (and silver) that is anywhere must be somewhere, the total quantity of precious metals to be held by all money-users was independent of the demand for it. The only way supply could be brought into conformity, and kept in conformity, with demand was through changes in the value of the commodity used as money.

[boldening: mine]

from Nicholas Kaldor wrote a major article in 1985 titled How Monetarism Failed (Challenge, Vol. 28, No. 2, link).

In his essay Keynesian Economics After Fifty Years, (in Keynes And The Modern World, ed. George David Norman Worswick and James Anthony Trevithick, Cambridge University Press, 1983), Kaldor wrote:

The excess supply would automatically be extinguished through the repayment of bank loans, or what comes to the same thing, through the purchase of income yielding financial assets from the banks.

Here’s the Google Books preview of the page:

In his article, Circuit And Coherent Stock-Flow Accounting, (in Money, Credit, and the Role of the State: Essays in Honour of Augusto Graziani, 2004. Google Books link) Marc Lavoie showed how this precisely works using Godley’s transactions flow matrix. (Paper available at UMKC’s course site). See Section 9.3.1

The fact that the supply of credit and demand for money appeared to be independent

 … has led some authors to claim that there could be a discrepancy between the amount or loans supplied by banks to firms and the amount or bank deposits demanded by households. This view of the money creation process is however erroneous. It omits the fact that while the credit supply process and the money-holding process are apparently independent, they actually are not, due to the constraints or coherent macroeconomic accounting. In other words, the decision by households to hold on to more or less money balances has an equivalent compensatory impact on the loans that remain outstanding on the production side.

So if households wish to hold more deposits, firms will have to borrow more from banks. If households wish to hold less deposits, they will purchase more equities (in the model) and hence firms will borrow lesser from banks and/or retire their debt toward banks.

Other References

  1. Lavoie, M. 1999. The Credit-Led Supply Of Deposits And The Demand For Money – Kaldor’s Reflux Mechanism As Previously Endorsed By Joan Robinson, Cambridge Journal Of Economics. (journal link)
  2. Kaldor N. and Trevithick J. 1981. A Keynesian Perspective On Money, Lloyd’s Bank Review.

Open Mouth Operations

In the previous post Is Paul Krugman A Verticalist?, I discussed the confusions economists and market commentators have on open market operations. Even top economists such as Krugman suffer from confusions on central banking and monetary matters.

I also mentioned the work of Alfred Eichner on bringing out more clarity on the defensive nature of open market operations. Let’s look at these matters more closely.

Before this let’s look at what people in general think. Most people think of open market operations as some kind of extra activity on the part of the central bank in collaboration with the bureau of engraving and printing and think of it as operational implementation of Milton Friedman’s helicopter drops! So when a central bank such as the Federal Reserve changes its target on interest rates – such as lowering the “Fed Funds target rate”, people start commenting as if the Federal Reserve is undertaking a mystical operation.

This is Monetarist or Verticalist intuition. It is easy to show that open market operations have nothing to do with fiscal policy and as we saw in the previous blog – very little with monetary policy itself. The open market operation of the central bank is not an income/expenditure flow such as government expenditure flows or tax flows and the former does not affect the net worth of the change the net worth of either the domestic private or the foreign sector. Hence it is hardly fiscal. Yet commentators and economists keep arguing that the central bank is “injecting money” into the economy!

Even Paul Krugman erred on some of these matters and was shown how to do good economics by Scott Fullwiler in his post Krugman’s Flashing Neon Sign. Missing everything Fullwiler was saying, Krugman wrote another post A Teachable Money Moment which has the following diagram:

Below we will see how Krugman’s Neoclassical/New Keynesian (whichever) intuition is flawed.

Setting Interest Rates

These matters (the public understanding) have become worse ever since the Federal Reserve and other central banks around the world started purchasing government debt in the open markets on a large scale – in programs called Large Scale Asset Purchases (also called “QE”). In the following I will consider cases when there is no “asset purchase program” by the central bank and tackle this issue later in another post.

A simple case to highlight how a central bank defends an interest rate (as opposed to changing it) is considering the corridor system. There target for overnight rates is usually at the midpoint of this “corridor”. At the lower end of the corridor banks get paid interest on their settlement balances they keep at the central bank while at the upper end it is the rate at which the central bank lends.

Because banks settle among each other on the books of the central bank, this gives the latter to fix the short-term rates and indirectly influence long term rates.

Why do banks need to settle with each other?

One of the first economists to understand the endogenous nature of money was Sir Dennis Holme Roberston who used to work for the Bank of England. In 1922, he wrote a nice book simply titled Money.

 From page 52:

. . . If A who banks at bank X pays a cheque for £10 to B who banks at bank Y, then Bank Y, when it gets the cheque from B, will present it for payment to Bank X: and bank X will meet its obligation by drawing a cheque for £10 on its chequery at the Bank of England. As a matter of fact, the stream of transactions of this nature between big banks is so large and steady in all directions that the banks are enabled to cancel most of them out by an institution called the clearing-house:  but the existence of these chequeries at the Bank of England facilitates the payment of any balance which it may not be possible at the moment to deal with in this way …

Because banks finally settle at the central bank finally, central banks have learned since their creation that they can set interest rates. This is strongest at the short end of the “yield curve” but directly or indirectly central banks also influence long term rates.

At the end of each day, some banks will be left with excess of settlement balances (if they see more inflows than outflows) and some banks will face the opposite situation. Because they need to satisfy a reserve requirement at the central bank (which can be zero as well), some banks would need to borrow funds from others. Borrowing means paying back with interest and this is where the central bank’s ability to target rates comes enters the picture.

For suppose some bank X needs funds and other banks wish to lend bank X at a very high interest rate. In this scenario bank X can simply borrow from the central bank, while other banks who demanded higher interest will see themselves with excess settlement balances earning less than the target rate. So it would have been better for the latter to have lent than keep excess balances. (Of course the qualifier is that these things are valid under scenarios when there is less stress on the financial system). Also with the same logic, the rate at which banks lend each other will not fall below the corridor because it will be foolish of a bank to have lent settlement balances to another bank when it could have earned higher by just keeping excess settlement balances at the central bank.

Here is a diagram from the post Corridors And Floors In Monetary Policy from FRBNY’s blog which explains central bank’s operations:

The other system as per this post is the floor system – in which the central bank’s target is the interest paid on settlement balances itself, rather than the midpoint of any corridor. This is what has been followed by the US Federal Reserve in recent times.

Back to the corridor system, an important question arises. Hopefully the reader is convinced that the overnight rate at which banks lend each other is between the corridor. However it is still unclear how and why the central bank can keep it at the midpoint.

If the central bank and the bankers understand the system well, it is possible for the central bank to be quite perfect in this. This happens for example in the case of Canada, where the bank is quite accurate in its objective.

The reason is that the central bank can easily add and subtract settlement balances by various means.

Take an example. Suppose the interest on reserves is 2.25%, the target is 3.00% and the discount rate – the rate at which the central bank lends overnight – 3.75%. If the central bank observes that banks are lending each other at 3.25% – slightly away from the target of 3.00%, it can simply create excess balances. Among the many ways, there are two –

  1. purchase/sale of government securities and/or repurchase/reverse repurchase agreements.
  2. shifts of government deposits from the central bank account into the Treasury’s account at banks or the opposite.

So the central bank knows how the curve in the figure above looks like and adds/subtracts settlement balances by the above methods (mainly). So banks are lending each other at 3.25%, the central bank will add settlement balances; if they are lending each other at 2.75% – the central bank will drain settlement balances.

More generally the “threat” by the central bank is reasonably sufficient to make banks lend at the target!

Open Mouth Operations

Let us suppose the central bank had been targeting 3.00% for three months now and decides to decrease rates.

What does it do?

Almost nothing!

The announcement itself will make banks gravitate toward the new target!

In his paper Monetary Base Endogeneity And The New Features Of The Asset-Based Canadian And American Monetary Systems, Marc Lavoie says:

When they [central banks] are not accommodating—that is, when they are pursuing “dynamic” operations as Victoria Chick (1977, p. 89) calls them—central banks will increase (or decrease) interest rates. As shown above, to do so, they now need to simply announce a new higher target overnight rate. The actual overnight rate will gravitate toward this new anchor within the day of the announcement. No open-market operation and no change whatsoever in the supply of high-powered money are required.

Hence the term “open mouth operations” which was coined by Julian Wright and Greame Guthrie in their paper Market-Implemented Monetary Policy with Open Mouth Operations.

Open Market Operations

The above paper by Marc Lavoie is an excellent source for open market operations and looking at central banking from an endogenous money viewpoint.

I mentioned in the previous post that the open market operations are defensive. In my analysis in this post till now, I ignored the factors which cause settlement balances of the banking sector as a whole to change. Let us bring in this complication.

Apart from banks, the Treasury – the domestic government’s fiscal arm – and other institutions such as foreign central banks, governments and international official institutions (such as the IMF) also keep an account at the (domestic) central bank. When these institutions transact, there is an addition/subtraction of banks’ settlement balances at the central banks.

Here’s one example. Suppose the Treasury transfers funds of $1m to a contractor for settlement of a project done by the latter. When the funds are transferred, the contractor’s bank account increases by $1m and the bank in which he banks sees its deposits and settlement balances rise by $1m while the central bank will reduce the Treasury’s deposits by $1m.

This may lead to a deviation of banks’ lending rate to each other and the central bank needs to drain reserves. The central bank can achieve this by shifting government funds at a bank into the central bank account. If the central bank transfers $1m of funds, banks’ deposits and settlement balances reduce by $1m and the Treasury’s account at the central bank will increase by $1m.

This is not an “open market operation” but another way of adding/draining reserves. In general, depending on institutional setups, the central bank may also do purchase/sale of government securities and/or repurchase agreements.

But this has nothing to do with policy itself – rather it is to maintain status quo. (Of course “large scale asset purchases” is a slightly different matter – first one needs to understand the corridor system).

In other words, this is a defensive behaviour on part of the central bank.

Alfred Eichner

In the previous post and in Alfred Eichner And Federal Reserve Operating Procedures, I mentioned about the contribution of Eichner. In an essay in honour, Alfred Eichner, Post Keynesians, And Money’s Endogeneity – Filling In The Horizontalist Black Box, (from the book Money And Macrodynamics – Alfred Eichner And Post-Keynesian Economics) Louis-Philippe Rochon says:

For Eichner, the overall or “primary objective [of the central bank], in conducting its open market operations, is to ensure the liquidity of the banking system,” which applies to either the accommodating or defensive roles. In either case, Eichner argues that “the Fed’s open market operations are largely an endogenous response to . . . the need both to offset the flows into and out of the domestic monetary-financial system and to provide banks with the reserves they require”; that is, resulting from the demand for money and the demand for credit respectively (1987, 847, 851).

And while the accommodating argument has been debated at length by post-Keynesians, the defensive role has been virtually ignored and only recently rediscovered (see Rochon 1999). Yet it is certainly Eichner’s greatest contribution to the post-Keynesian theory of endogenous money. . .

. . . The “defensive” behavior is defined by Eichner as the “component of the Fed’s open market operations [consisting] of buying or selling government securities so that, on net balance, it offsets these flows into or out of the monetary-financial system,” leaving the overall amount of reserves unchanged. This is the result of changes in portfolio decisions and increases or decreases in bank (demand) deposits. As a result of an increase in the nonbank’s desire to hold currency, for instance, “in order to maintain bank reserves at the same level, the Fed will need to purchase in the open market government securities equal in value to whatever additional currency the nonbank public has decided to hold” (Eichner 1987, 847).

In making the distinction between temporary and permanent open market operations, Rochon also quotes Scott Fullwiler:

Outright or permanent open market operations are primarily undertaken to offset the drain to Fed balances due to currency withdrawals by bank depositors. . . . Temporary open market operations are aimed at keeping the federal funds rate at its target on average through temporary additions to or subtractions from the quantity of Fed balances. Temporary operations attempt to offset changes in Fed balances due to daily or otherwise temporary fluctuations in the Treasury’s account, float, currency, and other parts of the Fed’s balance sheet, in as much as is necessary to meet bank’s demand for Fed balances. (2003, 857)

Paul Krugman

All this is completely opposite of Paul Krugman’s position that

. . . And currency is in limited supply — with the limit set by Fed decisions.

And Krugman’s mistake is not minor – it seems he is completely unaware of the huge difference money endogeneity makes.

So what is the difference between Krugman’s diagram and the one from FRBNY blog – even though they look similar? The difference is that the latter is descriptive of behaviour when policy is unchanged and is useful for describing open market operations etc while Krugman uses the same to describe policy changes – which in reality happen via open mouth operations. Paul Krugman confuses open market operations and open mouth operations. So much for a “teachable money moment”.

Krugman also shows his Monetarist intuition by claiming:

And which point on that curve it chooses has large implications for the economy as a whole. In particular, the Fed can always choke off a private-sector credit boom by moving up and to the left.

implying that the central bank in reality controls the monetary base and thence the money stock.

Some Post Keynesians argued since long ago that the central bank cannot control the money stock:

Here’s on Wynne Godley from from The Times, 16 June 1978:

(click to enlarge)